Continuity case · Timescape / Pantheon+
Later evidence matters — but it does not rewrite the 2007 claim
Wiltshire's 2007 proposal explains apparent cosmic acceleration within general relativity through regional differences in gravitational energy and clock rates. Later Timescape implementations were tested with Pantheon+. The question here is therefore not only how well the later model performs, but also what may legitimately be carried back to the historical claim.
Short result
The 2007 explanatory core is preserved in the later model lineage. The data and statistical interface, however, changed materially. Pantheon+ is therefore evidence about later Timescape implementations — not retroactive empirical confirmation or support of the historical GWB-T2 claim token.
Technical evidence details · Lane and Seifert
Lane · four distinct quantities: P+1690 intercept at z_min=0: ln B=2.93; P+580 curve peak: ln B=2.55; bootstrapped overall P+580 evidence: ln B=1.02±0.04 in favour of Timescape; integrated P+1690 evidence: marginally in favour of spatially flat ΛCDM.
Seifert · redshift profile: 0<z_min<0.023 strong to very strong preference; 0.023≤z_min<0.054 calibration regime with moderate to no significant preference; 0.054≤z_min<0.075 is not assigned a regime label in the source; from z_min≥0.075 the preference is exclusively moderate.
Seifert · full-sample comparison: ln B>5 for the full P+1690 sample in favour of Timescape over spatially flat ΛCDM. Additional BIC comparisons: ΛCDM with FLRW curvature ΔBIC=−13.39; wCDM ΔBIC=−11.70, both described by the authors as very strongly disfavoured relative to Timescape. ΔBIC and Bayes-factor ln B are not interchangeable.
Method qualifier: Seifert et al. state that full non-Gaussian modelling of the joint x₁,c distribution would require non-trivial changes to likelihood construction and integration; the Letter sidesteps this problem and leaves the full treatment for future work.
