Track D2 — Inverse burden: falsify "no distinguishing experimental test even in principle" (HYP-ENCODING-SCREEN)

Outcome: FAILED-TO-FALSIFY. The pillar survives its strongest 2024-2026 assault. Verdict-neutral; no registry movement; no new numeric ID. This is the 13th consecutive non-falsification of the standing verdict, and the second adversarial confirmation of Pillar 2 specifically (after the iter-6 distinguishing-test hunt).

What I tried (maximal-strength, assault mode)

I assumed the no-test corollary is WRONG and hunted the 2024-2026 phenomenology for a class-separating "reverse Weinberg-Witten" observable, steelmanning the single best candidate.

Best candidate (prong 1): the geontropic / Verlinde-Zurek interferometer signal (GQuEST). Its appeal as a break is real: the EFT proponents themselves concede the standard graviton-vacuum signal is ~10^-38 and "unobservably small," and that a detected ~10^-21 macroscopic length fluctuation "would signal a SEVERE BREAKDOWN of effective quantum field theory" (arXiv:2409.03894). So GQuEST is, on paper, a search for a signal the encoded EFT cannot produce — a literal candidate reverse-WW witness.

Best structural attack (prong 2): the global-content argument (Schneider 2024). The screen's factorization (A,ω)→E→{P(O)} rests on an unproven premise the wiki itself flags — that every UV/global/topological observable is a net invariant. Schneider argues global GR content is "irreducibly... not effective" and behaves like a superselection rule, attacking exactly that premise.

Why the attack collapses (honest assessment)

Prong 1 is a CLASS-vs-AXIS error. The geontropic signal separates EFT-violating from EFT-respecting quantum gravity — orthogonal to the wager (emergent-geometry vs fundamental-metric). The decisive, author-sourced fact: GQuEST states "diverse approaches to quantum gravity yield α=O(1)" — conformal field theory, dilaton theory, and hydrodynamics — and "predict metric fluctuations of the same scale." Dilaton/hydrodynamic models are not geometry-from-algebra-emergent, yet give the identical signal. A positive GQuEST therefore co-classes emergent with non-emergent. The amplitude (⟨φ²⟩∝α, α undetermined O(1)) re-imports the holographic η=1/4G at the modular→metric step rather than generating it. Screen-PASSING — a fresh published instance, not a counterexample.

Three further 2026 primary sources independently re-confirm the screen from new directions: the "before spacetime" QGEM paper (2602.06091) yields the same Newtonian phase in both pictures; modular-fluctuation cosmology (2503.04886) is a reinterpretation reproducing standard perturbations (and concedes the key mode's existence is open); gravitational non-Gaussianity (2606.09119) discriminates the gate (quantumness), explicitly "not its ontological status."

Prong 2 locates the seam but supplies no witness. Schneider sharpens where a discriminator could live (global/superselection content) — which is precisely the wiki's named opening — but "does not clearly establish" any observational distinction and ends on an explicit unresolved fork. This is the iteration-9 result from the philosophy side: the reverse-WW witness is strictly harder than the reverse-WW theorem, which is equivalent (Weiner's algebraic Haag theorem) to the OPEN carrier problem. A "there may be a seam" argument is a research program, not a referee-survivable in-principle discriminator. Per the anti-crank protocol, a plausibility hope is treated as a failed break.

Net value added (beyond non-falsification)

• The pillar's strongest 2026 empirical lever (GQuEST) is screen-passing on the proponents' own admission — the dilaton/hydrodynamics co-classing is a sharper, primary-sourced version of the iter-6 Sharmila-Vermeulen-Datta co-classing already on record.
• Schneider 2024 is an independent, externally-published localization of the screen's sole seam (global/topological/net content) exactly where the wiki already places it — corroboration, not breach.
• Confirms terminal saturation on Pillar 2: no in-principle discriminator was constructed in 2024-2026; the no-test corollary keeps its OP-46 MEDIUM-HIGH→HIGH grade and the net-naturality/carrier condition.

What WOULD have broken it (for the record)

A construction (not a hope) of an observable provably realizable only without fundamental metric degrees of freedom — i.e. an exhibited reverse-WW witness, which is strictly harder than the open carrier theorem. None exists or was sketched as of June 2026.

---

Track D2 referee — DEFEND Pillar 2 (no distinguishing experimental test even in principle) against its strongest assault

Outcome: ATTACK-FAILS-VERDICT-HOLDS. The finder's FAILED-TO-FALSIFY is upheld, honestly reached, and fully web-verified. Verdict unchanged (13th consecutive). Carrier count FIVE. Hedge unmoved. No registry movement. No new numeric ID.

Method

Read the binding state (CONCLUSION §4/§7/§8, HYPOTHESES HYP-ENCODING-SCREEN / HYP-FACTORIZATION-IS-GEOMETRY, the iter-11 terminal-saturation audit, the iter-13 how-close calibration). Default stance: REFUTE the finder — the break fails unless it exhibits a referee-survivable in-principle discriminator. Web-verified every load-bearing citation live, fetching primary sources (arXiv HTML where the PDF was FlateDecode-opaque or mis-cached).

Prong 1 (GQuEST/geontropic) — screen-passing, class-vs-axis error [verified on authors' own words]

The candidate's appeal is real and verbatim-verified: the EFT baseline (2409.03894) concedes its own ~10^-35 m vacuum-noise prediction is "unobservably small" and that a large detected variation "would signal a severe breakdown of effective quantum field theory." So GQuEST is, on paper, a hunt for a signal the encoded EFT cannot produce — a literal candidate reverse-WW witness.

It collapses on the wager's axis, and the decisive fact is the GQuEST authors' own (verbatim): "diverse approaches to quantum gravity yield α=𝒪(1). These include analyses from conformal field theory, dilaton theory, and hydrodynamics," which "predict metric fluctuations of the same scale." Dilaton/hydrodynamic/CFT models are NOT geometry-from-algebra-emergent in the wager's sense, yet give the identical signal — a positive GQuEST therefore co-classes emergent with non-emergent. And the amplitude is undetermined: ⟨φ²⟩∝α with α an O(1) parameter the authors "expect... will be calculated exactly" later — re-importing the holographic η=1/4G at the modular→metric step rather than generating it. This is a screen-PASSING instance, a primary-sourced version of the iter-6 co-classing and of CONCLUSION §7's recorded GQuEST read. A closed lever re-run, not a new decidable one. The proposed test co-classes; no generative construction is installed.

Three further 2026 primary sources, all verified verbatim, independently re-confirm the screen: (i) 2602.06091 — the most aggressively-titled "before spacetime" QGEM paper — concedes the familiar Newtonian 1/r phase "arises only after... introducing the infinity twistor," the same observable phase in both pictures; (ii) 2503.04886 (PRD 112 (2025) 026017) reproduces standard cosmological perturbations ("whenever an effective spherically symmetric perturbation is quantized... it can lead to the VZ scaling"); (iii) the non-Gaussianity line discriminates the gate (quantumness), not the ontological axis.

Prong 2 (Schneider global content) — locates the seam, supplies no witness [the serious attack; weighed and refuted]

Schneider's load-bearing claims are verified verbatim: global GR content is "irreducibly... not effective," "akin to a superselection rule," and "field observables in GR are only defined after... fixing the global manifold structure." This genuinely sharpens WHERE a discriminator could live — and it lands exactly on the sole seam HYP-ENCODING-SCREEN already names (the un-proven net-sharing assumption that every UV/global/topological observable is a net invariant). That is corroboration of the wiki's own opening, not a breach of it.

The refutation is that establishing the seam EXISTS is not exhibiting an observable that flows through it. Verified verbatim: the paper does NOT construct a discriminator ("distinguishing empirical content from theoretical content... is famously some trouble") and ends on an explicit unresolved fork ("Either... a topic of import within quantum gravity phenomenology... or else we ought to radically reorient ourselves"). This is the iter-9 result from the philosophy side: the reverse-WW WITNESS is strictly harder than the reverse-WW THEOREM, which is equivalent (Weiner, arXiv:1006.4726, verified verbatim in iter-11) to the OPEN carrier problem. A "there may be a seam" argument is a research program, not a referee-survivable in-principle discriminator — and the binding anti-crank protocol treats a plausibility hope as a failed break.

No-go / extraordinary-claim red-team — clean

No in-principle discriminator constructed. No generative-vs-encoding construction. No metric/symmetry smuggle. The proposed test co-classes emergent with fundamental on the authors' own admission. The seam-locating prong reproduces the wiki's named opening without supplying the missing witness. No conflict with iter-3 encode/generate, iter-6 co-classing, iter-7 η₀ vacuity, iter-9 naturality + reverse-WW=carrier, iter-10/11/12/13 closures. The extraordinary-claim gate does not fire.

Net

The best 2024-2026 empirical lever (GQuEST) is screen-passing on the proponents' own words; the best structural attack (Schneider) re-locates the screen's sole seam without breaching it. No referee-survivable in-principle discriminator exists or was sketched as of June 2026. The pillar survives its strongest assault. Positive value: three fresh web-verified 2026 screen instances, and an external philosophy-side localization of the seam exactly where the wiki places it. What WOULD have broken it (unchanged): an exhibited reverse-WW witness — an observable provably realizable only without fundamental metric degrees of freedom — which is strictly harder than the open carrier theorem. None exists.