Track R2 — The measurement problem, operationalized (2026-07-02)

Status: Research note (ROADMAP Tier 1, item 3). Deepens OP-2/OP-26/OP-27/OP-28, the A-3/QM-4 ledger entries (ASSUMPTIONS_LEDGER.md), GC-6, and the objective-collapse row of EXPERIMENT_WATCHLIST.md. Every reference below verified live 2026-07-02. No claim in this note bears on the carrier problem or the encoding screen; this is core/costs bookkeeping, not the wager.

1. The 2025–2026 objective-collapse bound state

Channel	Model	Bound (as verified)	Source	Tag
Underground spontaneous radiation (XENONnT SR1, 1–140 keV)	Markovian mass-proportional (white-noise) CSL	$\lambda/r_C^2 < 3.0\times10^{-3}\ \mathrm{s^{-1}m^{-2}}$ (90% CL); canonical GRW point ( $r_C=10^{-7}$ m, $\lambda=10^{-16}\ \mathrm{s^{-1}}$ ) deviates from the XENONnT best fit by 9.1σ — the first experimental exclusion of the originally proposed CSL values; ×135 over Majorana Demonstrator; most stringent for $r_C\lesssim10^{-5}$ m	arXiv:2506.05507; PRL 136, 120201 (2026)	ESTABLISHED
Underground spontaneous radiation (prior best)	same	Majorana Demonstrator	PRL 129, 080401 (2022)	ESTABLISHED
Underground spontaneous radiation	Markovian (radiating) Diósi–Penrose	$R_0 > 4.9\times10^{-10}$ m (90% CL), ~5× over prior; parameter-free DP already dead since $R_0>0.54$ Å (95%)	arXiv:2506.05507; Donadi et al., Nat. Phys. 17, 74 (2021), arXiv:2111.13490	ESTABLISHED
Space GW-detector testbed (LISA Pathfinder, rotational noise)	CSL	~2× improvement over translational bound for $r_C\sim10^{-5.5}$ – $10^{-3.5}$ m	arXiv:2501.08971; PRA 111, L020203 (2025)	ESTABLISHED
Space GW-detector testbed (LPF, updated acceleration data)	CSL / DP	$\lambda\lesssim8.3\times10^{-11}\ \mathrm{s^{-1}}$ at $r_C=10^{-7}$ m (mHz band); DP cutoff ≳285 fm	arXiv:2411.17588 [preprint; unverified journal-ref] (2017 baseline: $\lambda\lesssim2.96\times10^{-8}\ \mathrm{s^{-1}}$ , $R_0\gtrsim40$ fm, PRD 95, 084054 (2017))	INFERENCE, preprint rider
Matter-wave interferometry	any collapse (model-robust)	$>1.7\times10^5$ Da record, macroscopicity $\mu=15.5$ (already registered in the watchlist)	arXiv:2507.21211; Nature 649, 866 (2026)	ESTABLISHED
Theory input enabling the 2025–26 leap	CSL vs DP discrimination	low-energy X-ray emission is atomic-species- and model-dependent (electron–proton cancellation effects) — the radiation channel became model-discriminating	Piscicchia et al., PRL 132, 250203 (2024)	ESTABLISHED
Proposal (not a bound)	CSL, interferometric-class	reversible-nanoparticle-pointer test targeting $\lambda\sim1.4\times10^{-10}$ → $8.7\times10^{-12}\ \mathrm{s^{-1}}$	arXiv:2606.22707 (2026-06, preprint)	SPECULATIVE

Attribution correction. The 2506.05507 bound is the XENONnT Collaboration's (177 authors). VIP is a distinct program: VIP-3 is its Pauli-exclusion-principle arm (design: Entropy 26, 752 (2024); installing at LNGS, data-taking planned from 2025 per the design paper; latest VIP-2 PEP bound: violation probability $<\sim2\times10^{-43}$ , reported 2026-02); the collapse arm of that group is the HPGe/BEGe spontaneous-radiation program. ESTABLISHED

2. Surviving parameter windows INFERENCE synthesis on ESTABLISHED bounds

White-noise Markovian mass-proportional CSL: both historically motivated points — GRW (1986-canonical) and Adler (2007) — are excluded; the family is cornered between X-ray bounds ( $r_C\lesssim10^{-5}$ m) and GW-detector/LPF bounds (large $r_C$ ). What survives is parameter space with no independent motivation.
Radiating Markovian DP: survives only for $R_0>4.9$ Å — an atomic-scale smearing with no natural microscopic motivation.
Escape routes (explicitly flagged by XENONnT): colored-noise (non-Markovian) extensions — keV-radiation bounds evaporate by construction below the noise-frequency cutoff; dissipative extensions; non-mass-proportional couplings; non-radiating gravity-collapse variants. For all of these the binding constraints are mechanical (mHz–kHz) and interferometric, orders of magnitude weaker. Map anchor: Carlesso et al., Nat. Phys. 18, 243 (2022), arXiv:2203.04231. The class is not falsified; the parsimony cost of remaining in it has risen qualitatively — the models now survive by retreating from their original parameter targets. INFERENCE
The iteration-6 flag stands: "decisive coverage ~2035" remains an unanchored internal extrapolation; colored-noise cutoff space is unbounded below, so complete closure is not guaranteed by any funded roadmap. OPEN

3. Costs ledger, 2026 edition (updates the interpretation paragraph in domains/quantum-mechanics.md)

Program	Buys	Costs (2026)	2024–26 delta
Decoherence + branching (Everett)	universal unmodified unitarity; no new parameters; decoherence supplies branch structure	Born-weight account still open (OP-27/OP-28: incoherence + quantitative problems); branching ontology; preferred basis rests on decoherence approximations	no new no-go surfaced in this sweep OPEN; LF theorems (below) cost it nothing extra — it already rejects absoluteness of observed events
Objective collapse (GRW/CSL/DP)	single-world dynamics; falsifiability (unique among the options)	canonical parameter points now excluded (GRW 9.1σ; Adler long dead; parameter-free DP dead) → forced retreat to colored/dissipative variants: a new moving-target/parsimony cost; relativistic covariance unresolved; energy non-conservation tightly bounded	XENONnT PRL 136, 120201 (2026); LPF reanalyses; model-discriminating radiation theory (PRL 132, 250203 (2024))
de Broglie–Bohm pilot wave	single world, determinism, clean measurement account, explicit ontology	explicit nonlocality / preferred foliation; quantum-equilibrium hypothesis; relativistic/QFT awkwardness	no new no-go verified in this sweep OPEN; LF adds no new cost (local agency already rejected)
Relational QM	no absolute states; relational facts	live incompatibility charge CONTESTED: GHZ–Wigner-friend argument (Lawrence–Markiewicz–Żukowski, Quantum 7, 1015 (2023)) vs consistency defenses (Cavalcanti–Di Biagio–Rovelli, EJPS 13, 55 (2023); Drezet, Found. Phys. 54, 5 (2024)) and rejoinder (LMŻ, EJPS 15, 16 (2025), arXiv:2310.18008); the Adlam–Rovelli cross-perspective-links postulate restores intersubjectivity at the cost of departing from strict relationalism	exchange completed in print through 2025; unresolved
QBism	dissolves the measurement problem (state = agent's credence)	non-objective $\psi$ ; shared-reality/intersubjectivity burden	intersubjective-agreement analysis within QBism (arXiv:2312.07728) OPEN; LF costs nothing (absoluteness rejected)
Copenhagen/standard (single objective outcome + no dynamics change)	working physics as-is	now the directly targeted corner of extended-Wigner-friend theorems: retaining both absoluteness of observed events and local agency is LF-inequality-constrained territory	see below

Extended-Wigner-friend repricing ESTABLISHED theorems; CONTESTED significance. The Local Friendliness (LF) line hardened 2023–2026: "thoughtful" LF from strictly-weaker-than-Bell assumptions (Wiseman–Cavalcanti–Rieffel, Quantum 7, 1112 (2023); critique: Kent, arXiv:2302.12707); LF polytope proven equal to the Bell polytope in a wide class of multipartite/sequential EWF scenarios, with a systematic Kochen–Specker→LF translation (Walleghem–Yīng–Wagner–Schmid, Quantum 9, 1819 (2025)); "Noncontextual Friendliness" — absoluteness of observed events + noncontextual agency jointly inconsistent with QT (Walleghem–Catani, arXiv:2502.02461, rev. 2026-05, preprint); first LF-violation circuits on quantum simulators/hardware with scaling branch factor — proof-of-principle, loophole-ridden, error-limited (Zeng–Labib–Russo, Quantum 9, 1851 (2025)). Counterweight: the Schmid–Yīng–Leifer six-argument review (arXiv:2308.16220, preprint) argues EWF arguments hinge on assumptions about correlations inaccessible to any observer even in principle — independently vindicating GC-6's standing referee caution. Net: EWFS theorems constrain assumption combinations, never single interpretations; they reprice the ledger, they do not decide it.

4. Near-term decidable vs metaphysical INFERENCE

Decided (2021–2026): parameter-free DP — dead; canonical white-noise CSL points (GRW, Adler) — dead.
Being decided (now–2030s): the remaining Markovian CSL space (underground + LPF-class + levitated mechanics); colored/dissipative variants partially (mechanical bands; interferometric proposals approach the model-robust regime).
Programmatic (2030s+): LF violations with progressively more complex "friends" (quantum-computer ladder) — each rung eliminates assumption pairs at that complexity scale, never a specific interpretation.
Metaphysical as of 2026 (no distinguishing experiment known even in principle): Everett vs pilot-wave vs QBism vs RQM; Born-rule derivability (OP-27/OP-28); which LF assumption fails. The only lane where CONTESTED→decidable promotion is actually happening is objective collapse — and only its Markovian sub-family.

References (all verified live 2026-07-02)

XENONnT Collab., PRL 136, 120201 (2026) — https://arxiv.org/abs/2506.05507
Majorana Demonstrator, PRL 129, 080401 (2022) — https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.129.080401
Donadi et al., Nat. Phys. 17, 74 (2021) — https://www.nature.com/articles/s41567-020-1008-4 ; arXiv:2111.13490
Piscicchia et al., PRL 132, 250203 (2024) — https://link.aps.org/doi/10.1103/PhysRevLett.132.250203
Altamura–Vinante–Carlesso, PRA 111, L020203 (2025) — https://arxiv.org/abs/2501.08971
Dai–Miao–Ma, arXiv:2411.17588 [preprint] — https://arxiv.org/abs/2411.17588
Helou–Slagmolen–McClelland–Chen (2017 LPF baseline), PRD 95, 084054 (2017) — https://journals.aps.org/prd/abstract/10.1103/PhysRevD.95.084054
Carlesso et al., Nat. Phys. 18, 243 (2022) — https://arxiv.org/abs/2203.04231
Reversible-pointer CSL proposal, arXiv:2606.22707 [preprint] — https://arxiv.org/abs/2606.22707
VIP-3 design, Entropy 26, 752 (2024) — https://doi.org/10.3390/e26090752
Wiseman–Cavalcanti–Rieffel, Quantum 7, 1112 (2023) — https://quantum-journal.org/papers/q-2023-09-14-1112/
Kent, arXiv:2302.12707 [preprint] — https://arxiv.org/abs/2302.12707
Walleghem–Yīng–Wagner–Schmid, Quantum 9, 1819 (2025) — https://quantum-journal.org/papers/q-2025-07-31-1819/
Walleghem–Catani, arXiv:2502.02461 [preprint] — https://arxiv.org/abs/2502.02461
Zeng–Labib–Russo, Quantum 9, 1851 (2025) — https://quantum-journal.org/papers/q-2025-09-05-1851/
Schmid–Yīng–Leifer, arXiv:2308.16220 [preprint] — https://arxiv.org/abs/2308.16220
Lawrence–Markiewicz–Żukowski, Quantum 7, 1015 (2023) — https://quantum-journal.org/papers/q-2023-05-23-1015/
Cavalcanti–Di Biagio–Rovelli, EJPS 13, 55 (2023) — https://arxiv.org/abs/2305.07343
Drezet, Found. Phys. 54, 5 (2024) — https://arxiv.org/abs/2209.01237
Lawrence–Markiewicz–Żukowski, EJPS 15, 16 (2025) — https://arxiv.org/abs/2310.18008
QBism intersubjective agreement, arXiv:2312.07728 [preprint] — https://arxiv.org/pdf/2312.07728
SEP, Relational Quantum Mechanics — https://plato.stanford.edu/entries/qm-relational/