PUH Theorem 280 (Falsifiability Audit) — Every Near-Term Testable Prediction in the Archive, Tiered: Claim, Sharpness, Dataset Status, Kill-Condition; From Data-Exists-Now to Conditional; Postdictive Separated from Predictive; Where a Filed Prediction Lacks a Kill-Condition, the Audit Says So

FALSIFIABILITY AUDIT (T280). ZERO new claims (T276 survey pattern). Collects the near-term testable subset of the archive's 50+ falsifiable predictions; for each: claim + DOI, sharpness, dataset status, and KILL-CONDITION at filed strength. Where a filed claim lacks an explicit threshold, the audit says so rather than supplying one. Postdictive consistency separated from prediction. Four archive findings: kappa_tau Atlas flag RESOLVED (DERIVED-AS-FILED); T195 abstract/body discrepancy flagged (corrigendum recommended); one in-corpus DOI citation error noted (registry-canonical used); filed predictions lacking kill thresholds NAMED. Tags: PREDICTION / CONJECTURE / CONDITIONAL / POSTDICTIVE-CONSISTENT, all as filed. === DESCRIPTION (abstract) === Photonic Universe Hypothesis (PUH) — Falsifiability Audit. METHOD: an entry qualifies as near-term testable if a specified observable, in a named dataset or instrument class, can confirm or kill it within roughly the current decade. Each entry records claim (quoted at filed strength, with DOI), sharpness (parameter-free number / direction / conditional), dataset status (data exists now / funded 2026-2031 / dedicated experiment / gated), and kill-condition — the audit's spine. Where the filing supplies a threshold it is quoted; where it does not, the audit says "threshold not filed" (supplying one would be a new claim; none are made). Postdictive matches are quarantined under their own label. TIER 1 — TESTABLE NOW: (1) the Jovian coherent clock — a phase-stable 177.86-minute line in the existing 30-year Galileo+Juno JEDI record, with four filed discriminants (T161 + test-opportunity paper); kill: proper periodogram finds no such line. THE DATA EXISTS NOW. (2) The standing falsifier: gravitational waves are tensor-only (T193) — ANY confirmed scalar/vector mode falsifies; live at LIGO/Virgo/KAGRA, passing to date. (3) Directional clock anisotropy delta-nu/nu = (2/9)(r_S/r_E) = 3.1e-10, vertical vs horizontal co-located clocks — 400 billion times the demonstrated floor; kill: null at existing precision (T195/T185). (4) Clock-geodesy 11/9 = 1.222 offset vs seismic density; kill: agreement excluding a 22.2% systematic. TIER 2 — FUNDED 2026-2031: (5) kappa_tau = 16/15 = 1.0667 (T184) — the gold-standard entry: parameter-free chain (2/9)(18/60) = 1/15 (arithmetic verified; inputs trace to T152/T154 — the Correlation Atlas's UNVERIFIED flag is RESOLVED to DERIVED-AS-FILED), current 1.02 +/- 0.10, filed kill (below 1.040 at sigma < 0.013, 3 sigma) and confirm (above 1.050 at 2 sigma) thresholds, HL-LHC window 2028-2030, plus the joint generation-table pattern kill. (6) The Brookhaven four (Abidi = kappa_tau; Hock He-3 lattice drag; Khan incompressibility; Liu THz lattice modes) — directions specified, numeric thresholds NOT filed for Hock/Khan/Liu (said so). (7) CMB-S4 V-mode/LSS correlation (T190; registry-canonical DOI used; an in-corpus citation error assigning T189's DOI to T190 is noted). (8) The DESI w(z) branch arbiter (T252/T254): split (p=8) thawing vs compact (p=124) w = -1 exactly — adjudicates BETWEEN PUH's branches, not PUH wholesale. (9) The caustic P(k) excess (T159): P_PUH = P_CDM(1 + sigma_n^2 (k l_corr)^(2/3)) — 0.5%/0.7%/3.4% at k = 0.5/1/10 Mpc^-1 for sigma_n = 0.05, DESI/Euclid; bispectrum as the filed "key test"; plus P232.1 (T232) redshift-filament cross-correlation at 1e-3 to 1e-2 with filed kill verbatim. (10) Zero rocky-planet occurrence at z > 6 (T198); kill: ONE confirmed rocky planet. (11) The ~440 MeV scalar glueball (T189); tolerance not filed (said so). TIER 3 — DEDICATED EXPERIMENT: (12) 24-km balloon clock, 8.2e-22 (body derivation 3.4e-26/m; AUDIT FLAG: T195's abstract states 2.6e-24/m — internal discrepancy; corrigendum recommended). (13) Jupiter-orbit clock 8.7e-9 + 22% Juno-orbit GR excess. (14) Eris heat flux 1-4 mW/m^2 (T197). (15) Torsion-balance anisotropy (T185). (16) Horizon: sub-nm Casimir + Schwinger step (T170; DOI not re-verified — flagged). TIER 4 — CONDITIONAL/GATED: (17) VASCO anomalous flavour ratio (T234) — conditional on S1/S2/S3, whose S3 = the Leech elevation's Cor-Q2 (T279), gated on external Barbasch-Wong mathematics; NOT currently a live prediction, and if the gate fails no signature is predicted. (18) NANOGrav as cluster hum (T159) — CONJECTURE as filed. (19) Neuro-cosmos link — no discriminant filed; NOT-YET-TESTABLE. POSTDICTIVE-CONSISTENT (quarantined): Kagome PTI polarizations exactly 1/3 and 2/3 (T190; existing measurements); axion-insulator theta = pi, C_s = +/- 1/2 (Liu et al., Science 387, 162 (2026)); the Hubble ratio 1.083 (P178.2; matches the current 73.0/67.4 tension — predictive only for future shifts); delta_U = 2/27 matching Zenczykowski (PRD 2012-2013); Pluto 3.2 mW/m^2. ARCHIVE FINDINGS: (1) kappa_tau flag retired: DERIVED-AS-FILED. (2) T195 abstract/body discrepancy: corrigendum recommended. (3) T190 DOI citation error noted; registry-canonical used. (4) Missing kill thresholds named: Hock/Khan/Liu, CMB-S4 V-mode amplitude, glueball tolerance. NOT CLAIMED: any new prediction; any strengthening/weakening; that any test was performed here; that passing entries confirm PUH (consistent, not confirmatory); exhaustiveness (the archive states 50+ predictions; this is the near-term subset).