Distinguishing-Test Hunt (2026-06-10) — An In-Principle Test of the Wager: Found or Refused

Status: Targeted hunt, concluded. Verdict (B): no in-principle distinguishing test of the central wager exists in the 2024–2026 experimental/phenomenology literature, and none survives the net-invariant analysis. The mission's verified lead resolves negatively and becomes the cleanest published evidence FOR the encoding screen. Standing verdict unchanged — 8th confirmation; first via the experimental/phenomenology literature rather than analytic levers. Last updated: 2026-06-10 Tracks: H1 (literature sweep), H2 (reverse-Weinberg–Witten / net-invariant analysis), H3 (deep-read of the verified lead). All three adversarially refereed; corrections binding over original claims.

This note attacks CONCLUSION §7's sharpest sentence — the central wager has no distinguishing experimental test even in principle — from the side the analytic iterations (4–7) could not reach: the actual 2024–2026 experimental and phenomenology literature, plus the one theorem-shaped escape (a reverse Weinberg–Witten). If any referee-upheld candidate had distinguished-in-principle, this would have been clause/verdict-level: §7 would need re-wording with progress-toward-physics hedges. None did. The sentence stands and gains positive evidence.

1. What "distinguishing" requires, and the null taken seriously

The wager (W): in our $\Lambda>0$ universe the fundamental description is non-geometric — a net of von Neumann algebras with states $(\mathcal A,\omega)$ , equivalently a causal order plus modular/entanglement data — and the Lorentzian metric with its Einstein dynamics is derived by a readout $\Phi$ . The rival (G): metric geometry (possibly quantized/discrete) is fundamental; the algebra is derived. Both reproduce the cores (Hilbert-space QM with exact linearity, the SM, GR at tested precision, ΛCDM).

A class-level distinguishing test (the only standard that "tests the wager" under the anti-crank protocol) is an observable $O$ , outcome set $S$ with $\inf_{T\in\mathcal T_A}P_T(O\in S)>\sup_{T\in\mathcal T_G}P_T(O\in S)$ (or $A\leftrightarrow G$ ). Member-level separation is trivial; class separation is the bar. The null, taken seriously: the wager is a claim about derivational order, which is not an observable, so any observable feature of the prior structure can in principle be re-housed as a feature of a fundamental-geometry theory — unless a feature provably cannot be re-housed. The hunt's burden was to find that feature, or to show why none currently can be found.

2. H1 — the 2024–2026 literature sweep

The verified lead resolves negatively. Sharmila–Vermeulen–Datta (Nat. Commun. 17, 701 (2026); online 23 Dec 2025; arXiv:2505.22892) give a three-class taxonomy of spacetime-fluctuation two-point functions:

class (a) factorised — Oppenheim CQ (fundamental classical metric), CSL, Diósi–Penrose;
class (b) inverse/polynomial — Károlyházy (fundamental-metric stochastic), Verlinde–Zurek geontropic (emergent), EFTs;
class (c) exponential — entanglement-holographic / mesoscopic.

The decisive fact, verified verbatim against the body text: "Our approach is agnostic to the microscopic origins of the SFs." The classification variable is the two-point correlator $\rho$ — phenomenological and mechanism-blind. The model→class map is therefore many-to-one: geontropic (emergent) and Károlyházy (fundamental-metric) BOTH sit in class (b); Oppenheim CQ and CSL BOTH sit in class (a). No class is realizable only by a non-fundamental-metric model. The taxonomy distinguishes gravity models, never fundamental-vs-emergent ontology; the authors' own ceiling is "evidence for or against one of the above classes of correlation functions." This is the cleanest concrete instantiation to date of HYP-ENCODING-SCREEN Corollary 2 (UV-portability) — a published, peer-reviewed worked example of the screen, not a counterexample. It belongs in the GQuEST watchlist row as positive evidence. INFERENCE, high — chain audited link-by-link, §4

Two author-side concessions of observational degeneracy.

Sidan A & Banks (arXiv:2502.15108). Holographic-spacetime inflation: predictions "fit extant data equally well" (verbatim) as field-theoretic inflation, differing only conceptually ("radically different", no transplanckian problem) — not observationally. Quotation-hygiene correction (binding): the sentence "No claim is made that the CMB could observationally distinguish emergent/holographic spacetime from conventional metric inflation" is the hunt's paraphrase, not a verbatim Banks quote — recorded as such (EPISTEMICS §4). The degeneracy reading is accurate; only the quotation hygiene was at fault.
Erlich (arXiv:2504.03084). Stochastic composite gravity: low-ℓ CMB modification "not unique"; predictions "could emerge from either composite or fundamental theories." Nuance on record: Erlich frames an out-of-equilibrium stochastic-vs-QFT disagreement as a positive test — but it bears on exact linearity of QM (the gate / objective-collapse axis), NOT geometry-priority, so it is not a reverse-WW.

The rest of the sweep, screened. Schützhold's ~10⁶ km graviton-manipulation interferometer (PRL 2025, HZDR — verified) tests the radiative-sector gate (do gravitons exist), self-stating it "would not directly prove the existence of gravitons." Aalsma–Bak (arXiv:2503.04886) reproduce the standard CMB spectrum by quantizing an s-wave mode by hand — "whether such a mode exists ... remains an open question" (verbatim). Calcagni–Modesto (arXiv:2206.07066, JHEP 12 (2024) 024) give a member-level $r_{0.05}\approx0.01$ + blue SGWB from one UV-complete construction — also producible by fundamental-metric quantized gravity. Hashimoto et al. (arXiv:2411.16052) train a neural net assuming AdS/CFT — emergence presupposed, the encode-only architecture made literal. The Google traversable-wormhole simulation (Nature 612, 51 (2022)) + Kobrin–Schuster–Yao criticism (arXiv:2302.07897): the dual is a dictionary over an engineered system, not a claim about our universe. Analog gravity is dead by construction (the "spacetime" is an emergent description of a known fundamental substrate, the condensate). Net tally: 1 lead sharpened, 2 program-adjacent imprints, 1 gate, 4 dead, 1 member-level, 1 OPEN theorem target. Nothing flips not-yet-physics to physics.

3. H2 — reverse-Weinberg–Witten and the net-invariant closure

The theorem-shaped object that would void the screen: an observable $X$ , outcome set $S$ , realizable with no fundamental metric d.o.f. and by no core-satisfying fundamental-metric theory — a reverse Weinberg–Witten (WW itself runs the other way, forbidding covariant- $T_{\mu\nu}$ composite massless spin-2, evaded by holography / vanishing $T_{\mu\nu}$ ; the evasion is symmetric across both classes — Carone–Claringbold–Vaman, arXiv:1710.09367, verified).

Every candidate UV/global/topological/boundary discriminator turns out to be an invariant of the operator-algebra net, hence EFT-shared:

theta-angle / global form / line-operator spectrum — a DHR / Haag-duality-violation invariant, now classified for higher-form symmetries (Casini–Magán, arXiv:2511.21810, verified live);
't Hooft anomaly — IR-matched (part of $E$ 's observable content) and a net invariant: doubly screened;
soft hair / memory — the IR sector of the same graviton EFT (soft theorems = Ward identities), semiclassical on a background;
topological entanglement entropy $\gamma=\log\mathcal D$ — the total quantum dimension of the DHR superselection category, a net+state invariant; diagnoses topological order, not derivational order (Parnachev–Poovuttikul, arXiv:1504.08244);
modular flow / modular Berry phase / kinematic-space symplectic form — Tomita–Takesaki-built net+state invariants (Czech et al., arXiv:2305.16384; Huang–Ma, arXiv:2110.08703); and $J$ structurally carries reflection positivity (iter-7 A1-F4), so it sits on the Euclidean side and cannot be a secret Lorentzian discriminator;
indefinite causal order — reduces operationally to definite acyclic order (Vilasini–Renner, PRA 110, 022227 (2024)).

Binding correction (governs the H2 headline). The submission framed this as "a reverse-WW theorem is structurally impossible." That overclaims, exactly as the iteration-6 referee already struck (R6-F7, binding). "No reverse-WW theorem exists" is a literature-absence claim, tag OPEN, as of 2026-06, NOT a structural impossibility. The impossibility argument ("a witness $X$ would depend on derivational order, which is not a net invariant; every UV/global observable IS a net invariant") only restates the screen under the net-sharing assumption A1 (both classes realize the same net isomorphism class). A1 is the load-bearing, unproven premise — and is itself the reopening lever. A no-global-symmetry result under A1/holography (SymTFT-inside-holography: Heckman–Hübner–Murdia, PRD 110, 046007 (2024), arXiv:2401.09538) does not upgrade absence to impossibility. The reverse-WW target is therefore: dead as a route to a constructible test now, but OPEN as a theorem.

Registry-integrity correction (binding, AGENTS §3 / EPISTEMICS §4). Reverse-WW is NOT "OP-50." In the canonical wiki OP-50 is "the closed-universe Hilbert space: is it one-dimensional, and is QM only observer-relative there?" CONTESTED (OPEN_PROBLEMS.md, registered iteration 6). Reverse-WW was only proposed as OP-50 in an iteration-6 working note; the ID was assigned elsewhere and reverse-WW was never registered. It lives un-IDed inside HYP-ENCODING-SCREEN's portability corollary, tag OPEN, as of 2026-06. Every "OP-50 = reverse-WW" label is struck.

4. H3 — the verified lead, deep-read (why it sharpens rather than threatens the screen)

Claim under test: does the exponential/power-law/factorised correlator taxonomy distinguish emergent-metric from fundamental-metric-plus-noise? The chain: classification variable = the two-point correlator $\rho$ (phenomenological) → class is a functional of $\rho$ alone → two theories with identical $\rho$ are co-classed by definition → the paper's model table is many-to-one (CSL, Diósi–Penrose, Oppenheim, Károlyházy → class (a)) → the model→class map is non-injective, hence non-invertible → mechanism (metric fundamental vs emergent) is unrecoverable from class → no class is emergent-only (every kernel is reproducible by classical stochastic metric noise; emergent models merely co-inhabit class (c)) → no reverse-WW observable is supplied → does not distinguish the wager. The authors' own ceiling — "evidence for or against one of the above classes of correlation functions" — sits exactly on the shared EFT/correlator layer of HYP-ENCODING-SCREEN. The taxonomy is a textbook instance of the screen, not a counterexample to it. [Track H3 unrefereed; its chain is fully subsumed by the H1 referee's UPHELD verdict on the same candidate, citation-audited at the strongest point.]

5. The screening argument, restated (verdict B)

Proposition (ENCODING-SCREEN, unchanged). INFERENCE — near-definitional given the encode-only classification If $\Phi$ is encode-only (consumes background data $g_1$ – $g_7$ , outputs no metric data beyond input, $\Delta_{\rm geo}\le0$ ) and a wager-true theory's entire bridge passes through $\Phi$ , then its prediction map factors $(\mathcal A,\omega)\to E\to\{P(O)\}$ through the semiclassical EFT $E$ whose background $\Phi$ consumed, so no observable separates it from any geometry-first theory with IR limit $E$ . The physics is the readout-independent classification fact (iters 3–7) that every current Program-A construction is encode-only.

Corollary (one-way, unchanged). A class-level test exists only if some wager-true theory has a non-factoring (generative) bridge; sufficiency is OPEN. The OP-46/§8 verdict-flipper (now compressed to the carrier problem, HYP-FACTORIZATION-IS-GEOMETRY) and the missing §7 experiment coincide as obstructions, not as guarantees.

What this hunt adds. Not a new hedge and not a removed one — positive evidence. Before today the portability corollary was a survey-scoped inference. It is now instantiated by (i) a peer-reviewed mechanism-blind taxonomy that co-classes emergent and fundamental models, and (ii) two emergent-spacetime programs conceding observational degeneracy in their own words. The screen's grade (MEDIUM-HIGH, conditional on the modular-constructibility ansatz, inherited from OP-46) and the §7 no-test corollary are unchanged.

6. Verdict

(B). No candidate distinguishing test exists in the 2024–2026 experimental/phenomenology literature, and none survives the net-invariant analysis. Every referee-upheld candidate is program-adjacent (SVD taxonomy, GQuEST-VZ, Aalsma–Bak), gate-core-target (Schützhold; the anomaly/soft-hair/TEE/modular-Berry net invariants), member-level (Calcagni–Modesto), or dead (Banks CMB, Hashimoto ML, wormhole/SYK simulation, analog gravity). The verified lead resolves negatively and becomes the screen's best published evidence. The theorem-shaped escape — a reverse Weinberg–Witten observable — remains unattempted and OPEN, as of 2026-06; it is not a proven structural impossibility, and it is not OP-50. CONCLUSION §7 stands as written, buttressed, with no hedge movement. 8th confirmation.

References

All verified live this session unless bracketed; two journal-coordinate riders flagged unverified in BIBLIOGRAPHY.

B. Sharmila, S. M. Vermeulen, A. Datta, Signatures of correlation of spacetime fluctuations in laser interferometers, Nat. Commun. 17, 701 (2026), DOI 10.1038/s41467-025-67313-3, arXiv:2505.22892. [load-bearing quote verified verbatim]
Sidan A, T. Banks, holographic-spacetime inflation / CMB, arXiv:2502.15108 (2025). [degeneracy quotes verified; one paraphrase de-quoted]
J. Erlich, stochastic composite gravity (GRF 2025 essay), arXiv:2504.03084 (2025).
R. Schützhold, Stimulated Emission or Absorption of Gravitons by Light, Phys. Rev. Lett. (2025), HZDR. [PRL coordinate unverified this session]
L. Aalsma, S.-E. Bak, Modular Fluctuations in Cosmology, arXiv:2503.04886; PRD 112, 026017 (2025). [arXiv identity + quote verified]
G. Calcagni, L. Modesto, primordial GWs from Weyl-invariant QG, arXiv:2206.07066; JHEP 12 (2024) 024.
K. Hashimoto, K. Matsuo, K. Murata, G. Ogiwara, D. Takeda, ML emergent-spacetime from linear response, arXiv:2411.16052.
D. Jafferis et al., traversable-wormhole simulation, Nature 612, 51 (2022), DOI 10.1038/s41586-022-05424-3; G. Kobrin, T. Schuster, N. Y. Yao (criticism), arXiv:2302.07897.
C. Casini, J. M. Magán, A generalization of the DHR theorem for higher form symmetries, arXiv:2511.21810 (2025). [verified live]
C. D. Carone, T. V. B. Claringbold, D. Vaman, Composite graviton self-interactions in a model of emergent gravity, arXiv:1710.09367. [authors corrected]
J. J. Heckman, M. Hübner, C. Murdia, On the Holographic Dual of a Topological Symmetry Operator, PRD 110, 046007 (2024), arXiv:2401.09538.
A. Parnachev, N. Poovuttikul, Topological Entanglement Entropy, Ground State Degeneracy and Holography, arXiv:1504.08244 (2015).
B. Czech, J. de Boer, D. Espíndola, B. Najian, J. van der Heijden, J. Zukowski, Changing states in holography: From modular Berry curvature to the bulk symplectic form, arXiv:2305.16384; W.-z. Huang, M.-z. Ma, Emergence of Kinematic Space from Quantum Modular Geometric Tensor, arXiv:2110.08703.
V. Vilasini, R. Renner, PRA 110, 022227 (2024); companion arXiv:2408.13387. [already on record]

Referee verdicts (per track)

Referee verdict — H1-literature-sweep (adversarial)

Stance: REFUTE by default; every load-bearing citation web-checked live 2026-06-10, every "follows-from-wager" link re-checked, the reverse-Weinberg–Witten statement checked against Weinberg–Witten, Coleman–Mandula, and HYP-ENCODING-SCREEN. Net: 10/10 candidates KEPT with verdicts UPHELD; one MAJOR (registry-ID misattribution of "OP-50"); three MINOR (a paraphrase dressed as a Banks quote; a date/volume slip; an author-less WW-evasion citation); the standing CONCLUSION verdict is unchanged for the 7th time and mildly strengthened — in the corrected form.

Citation audit (live web)

arXiv:2505.22892 (the VERIFIED LEAD) — authors (Sharmila, Vermeulen, Datta), title, and three-class taxonomy CONFIRMED; the load-bearing quote "Our approach is agnostic to the microscopic origins of the SFs." CONFIRMED verbatim in body; class assignments CONFIRMED (geontropic + Károlyházy both in class b; Oppenheim CQ + CSL both in class a). The resolution "no class is realizable only by a non-fundamental-metric model" is source-grounded. Bibliographic fix: Nat. Commun. 17 (2026) (online 23 Dec 2025), not "2025".
arXiv:2502.15108 (Banks) — title/authors and "fit extant data equally well" / conceptual-not-observational CONFIRMED. MINOR defect: the sentence the submission quotes ("No claim is made that the CMB could observationally distinguish…") is NOT in the paper — it is a paraphrase mis-presented as a verbatim quote. Demote to paraphrase. Verdict ("dead") stands.
arXiv:2503.04886 (Aalsma–Bak), arXiv:2206.07066 (Calcagni–Modesto JHEP 12(2024)024), arXiv:2411.16052 (Hashimoto et al.), arXiv:2302.07897 (Kobrin–Schuster–Yao), Nature 612, 51 (Jafferis et al.) — all resolve to claimed authors/titles; the Aalsma–Bak open-question quote and the Hashimoto "automatically emergent as its interpretable weights" line CONFIRMED verbatim.
Schützhold — UPGRADE unverified→[verified]: "Stimulated Emission or Absorption of Gravitons by Light," PRL 2025 (HZDR). Gate verdict corroborated.
arXiv:1710.09367 — resolves to Carone–Claringbold–Vaman, "Composite graviton self-interactions in a model of emergent gravity"; the vanishing-T_μν WW-evasion physics is correct, but the submission's citation gives no authors and an imprecise label — fix the reference.
Erlich arXiv:2504.03084 — confirmed GRF-2025 essay; paper itself calls the low-ℓ effect "not unique"; the out-of-equilibrium stochastic-vs-QFT disagreement is framed as a positive test of exact linearity (gate axis), not of geometry-priority — so still no reverse-WW. Portability verdict holds.

Chain re-check (is each link entailed by the wager?)

Every candidate's "distinguishing" link collapses exactly where the submission says it does: GQuEST re-imports η=1/4G at C2; the SF taxonomy is mechanism-agnostic input; Hashimoto presupposes AdS/CFT; the wormhole/SYK dual is a dictionary; analog gravity sits on a known substrate; Banks and Erlich concede degeneracy. No link is entailed by the wager that the submission claims to be. The reverse-WW statement is the correct theorem-shape, consistent with Weinberg–Witten (which runs the opposite way and is evaded by holography/vanishing-T_μν) and with HYP-ENCODING-SCREEN's portability corollary; Coleman–Mandula is not load-bearing and is not misused. The reverse-WW remains a literature-absence claim correctly tagged OPEN, as of 2026-06.

NO-GO red-team / the one MAJOR

The submission repeatedly brands the reverse-WW target "OP-50." This is a registry-integrity defect: canonical OP-50 is the closed-universe Hilbert-space problem CONTESTED (OPEN_PROBLEMS.md), and the reverse-WW item — although proposed as OP-50 in the iteration-6 working note — was never registered; it lives un-IDed inside HYP-ENCODING-SCREEN. The "OP-50 = reverse-WW" label is dead/superseded and must be struck wherever the submission uses it; refer to the target as "the reverse-Weinberg–Witten statement (HYP-ENCODING-SCREEN portability corollary), un-registered, OPEN." This does not touch the physics verdict.

Verdict-bearing summary

The headline — no in-principle distinguishing test was produced in 2024–2026; the VERIFIED LEAD resolves negatively because the SF taxonomy is mechanism-agnostic and co-classes fundamental and emergent models — is UPHELD, with the load-bearing primary-source quotations verified verbatim. Two author-side degeneracy concessions (Banks; the Hashimoto encode-only architecture) and the SVD shared-class taxonomy are genuine fresh confirmations of the ENCODING-SCREEN portability corollary, and the watchlist GQuEST row should indeed cite arXiv:2505.22892. The standing verdict is unchanged (7th) and no new analytic hedge is required. Binding corrections: (1) strike "OP-50" for reverse-WW everywhere [MAJOR]; (2) demote the Banks "no claim is made…" string from quote to paraphrase [MINOR]; (3) "Nat. Commun. 2025" → "Nat. Commun. 17 (2026)" [MINOR]; (4) add authors (Carone–Claringbold–Vaman) to the 1710.09367 citation [MINOR]; (5) mark Schützhold [verified]. With these applied, all ten candidate verdicts hold.

Referee verdict — H2-reverse-weinberg-witten (adversarial, 2026-06-10)

Stance: REFUTE, default-on. Every load-bearing citation web-checked live this session; the submission's structural claims checked against the binding iteration-6 referee correction (R6-F7), the iteration-7 A1 finding (η₀/J-reflection-positivity), and the actual registry IDs. Net: verdict UPHELD and tightened; the headline overclaim demoted; one registry-ID conflation corrected. 7/7 candidates kept with corrected labels; one MAJOR (the RWW-impossibility framing) and one MAJOR registry error (OP-50 conflation); no fabrication.

Citation audit — clean. All seven load-bearing arXiv IDs resolve to real papers that support their textual claims: Casini–Magán 2511.21810 (DHR→higher-form, order/disorder ↔ HDV net data — the submission's keystone, verified verbatim against the abstract); Vilasini–Renner 2203.11245 (PRA 110, 022227 (2024)); Parnachev–Poovuttikul 1504.08244 (TEE↔GSD↔holography); Czech/de Boer et al. 2305.16384; Huang–Ma 2110.08703; Heckman–Hübner–Murdia 2401.09538 (PRD 110, 046007 (2024)); Carney–Karydas–Sivaramakrishnan 2409.03894 (PRD 113, 106002 (2026)). No invented citation, number, or theorem. Two minor bibliographic notes: (i) the PRL 133, 080201 (2024) companion to Vilasini–Renner is cited under the PRA's arXiv ID — the ID resolves to the PRA paper; (ii) 2305.16384 is first-authored by Czech, not de Boer ("de Boer et al." is acceptable shorthand). Neither affects a verdict.

R-H2-F1 — The "net invariant ⇒ EFT-shared" mechanism. KEEP — severity MINOR (genuine sharpening). Recasting each candidate discriminator as a net (or net+state) invariant — DHR/HDV superselection data, IR-matched anomalies, total quantum dimension, Tomita–Takesaki modular data — is correct, citation-supported, and a legitimate unification of the iter-6/7 case-by-case survey into one checkable property. This is the submission's real contribution and it survives. But the mechanism is load-bearing only under assumption A1 (T_A and T_G realize the SAME isomorphism class of local-algebra net). A1 is what makes the screen near-definitional; it is stated honestly in the submission's own assumptions block, but the headline ("the loophole-class is CLOSED by a structural reason") oversells a result that is conditional on A1 + the modular-constructibility ansatz + encode-only. Grade it INFERENCE, MEDIUM-HIGH conditional, exactly as the standing HYP-ENCODING-SCREEN already carries — no hardening beyond the registry is earned.

R-H2-F2 — "Reverse-Weinberg–Witten is structurally impossible." CORRECTED — severity MAJOR (binding). This is the headline, and it re-imports the theorem-clothes the house already stripped in R6-F7 (binding): "'No reverse-WW theorem exists' is a literature-absence claim — tag OPEN, as of 2026-06." The submission's impossibility argument — a witness X would depend on derivational order, which is not a net invariant — merely restates the screen under A1; it does not prove no class-separating observable exists, because the whole question is whether A1 holds (i.e. whether some wager-true theory has a generative, non-net-invariant bridge). Per the standing one-way corollary, that generative bridge is necessary, not sufficient, and sufficiency is OPEN. So "impossible" is unsupported in both directions. The no-global-symmetry-in-QG and SymTFT-inside-holography results (2401.09538, verified) are themselves A1/holographic-duality statements and cannot lift absence to impossibility. Binding correction: demote "a reverse-Weinberg–Witten theorem is impossible" / "structural obstruction to its existence" to "no such theorem is known and none has been constructed; the screen is a near-no-go conditional on A1, the modular-constructibility ansatz, and the encode-only classification." WW itself is characterized correctly.

R-H2-F3 — Registry-ID conflation: "OP-46/OP-50 reopening condition … they coincide." CORRECTED — severity MAJOR (factual). The submission ties the screen's reopening condition to OP-50 ("the OP-46/OP-50 reopening condition — they coincide, as the one-way corollary already records"). In the live registry, OP-50 is the closed-universe Hilbert-space problem (is it one-dimensional / is QM only observer-relative there), tagged [CONTESTED], three-cornered (Harlow–Usatyuk–Zhao / Nomura–Ugajin / Miyata et al. 2606.04575). The "reverse Weinberg–Witten" was only proposed as OP-50 in the iter-6 distinguishing-prediction note; the ID was assigned to the closed-universe problem instead, and reverse-WW was never registered as an OP. The actual coincidence the registry records is OP-46 / HYP-FACTORIZATION-IS-GEOMETRY (the carrier problem) — the fifth-converging-route closure of iteration 7. Replace every "OP-50" in the submission's screen discussion with "HYP-FACTORIZATION-IS-GEOMETRY (the carrier problem)"; do NOT register a new HYP-SCREEN-TIGHT or re-use OP-50 for reverse-WW.

R-H2-F4 — "J carries reflection positivity ⇒ no modular observable is secretly Lorentzian-discriminating." KEEP — severity NONE. Correctly invokes the iteration-7 A1-F4 result (⟨a*Ω, JaΩ⟩ ≥ 0 by natural-cone self-duality; severity-NONE, independently numerically replicated). The modular conjugation sits on the Euclidean/positive side; this genuinely closes the modular-Berry candidate against a hidden Lorentzian discriminator. The single strongest link in the submission.

R-H2-F5 — Candidate verdict labels. KEEP — all correct. program-adjacent-not-wager (theta, TEE, modular, ICO) and gate-core-target-only (anomaly, soft hair) are the right buckets; none is dressed as a wager-discriminator, and no fabricated discriminator was offered (which would have been disqualifying). The reverse-WW candidate's "dead" label is kept for the route but its impossibility framing is corrected per R-H2-F2.

Red-team of the screen itself. (i) "Too definitional" — conceded by the submission and by the house; the value is unification + a precise reopening condition, not a theorem. (ii) The new Casini–Magán input genuinely strengthens the theta/anomaly arm (the higher-form case is now literature-classified as net data), which is a real 2025 advance over iter-6 — credit where due. (iii) The submission does not manufacture a test and does not move N*, the experiment watchlist, or any closed item — NOT-YET-PHYSICS stands.

Verdict-bearing summary. Standing verdict UNCHANGED (seventh consecutive iteration) and modestly tightened in corrected form: the global/topological/boundary discriminators are net(+state) invariants and therefore EFT-shared conditional on A1, sharpening — not replacing — HYP-ENCODING-SCREEN at its existing MEDIUM-HIGH (ansatz-conditional) grade. Two binding corrections govern the body: (1) the reverse-Weinberg–Witten impossibility claim reverts to a literature-absence OPEN per R6-F7; (2) the OP-50 reference is a registry error — the reopening condition coincides with HYP-FACTORIZATION-IS-GEOMETRY (the carrier problem), not OP-50 (closed-universe). No new registry ID (HYP-SCREEN-TIGHT) is warranted beyond a refinement note on HYP-ENCODING-SCREEN. No fabrication; citations clean.

Relevant files: C:/PROJECTS/FAFO/universal-physics/HYPOTHESES.md (HYP-ENCODING-SCREEN, HYP-FACTORIZATION-IS-GEOMETRY, HYP-MODULAR-TRICHOTOMY — lines 345, 357, 355); C:/PROJECTS/FAFO/universal-physics/OPEN_PROBLEMS.md (OP-50 actual definition lines 548–559; iter-7 progress line 581); C:/PROJECTS/FAFO/universal-physics/notes/2026-06-10-iter6-distinguishing-prediction-hunt.md (R6-F7 binding correction, lines 30, body §4); C:/PROJECTS/FAFO/universal-physics/notes/2026-06-10-iter7-indefinite-pairing-attempt.md (A1-F4 J-reflection-positivity).