Pantheon+ delivers H₀ = 73.04 +/- 1.04 from 1,701 calibrated supernovae, 5σ from Planck, surviving every documented cut, and its residuals carry structure the model forbids: host-environment correlations beyond the mass step, coherent low-z flow features, and w drifting under sample partitions. The compilations disagree at the margins in exactly the environment-sensitive regime, and each correlation must be processed as a systematic because the model has no place for it as physics.
The supernova Hubble diagram must be one smooth function: a single global expansion, every sightline statistically identical after standard corrections. Environmental dependence of inferred distances is impossible by construction, so the assumption converts any genuine environmental signal into an endless audit of systematics.
SCT predicts exactly the leakage Pantheon+ keeps showing. The supernovae sample a genuinely inhomogeneous Λ_eff field (P17, P19): calibrators inside the KBC volume inherit the local enhancement that sets the headline 73, the same environmental physics as the megamaser and ladder values, while the frame-tree corrections of hereditary time (P10) add systematic redshift offsets of 10⁻⁴ to 10⁻⁵ in (1+z) that differ between cluster-environment and field-environment hosts. The registered structure is specific: cluster-hosted supernovae should yield H₀ offset from field-hosted ones by 0.5 to 1 percent, the apparent w should drift when an inhomogeneous Λ_eff is forced into a homogeneous fit (the DESI artifact, derived at w₀ = -0.94, w_a = -0.58), and Hubble residuals should correlate with host environment after every standard correction, which is the structure the analyses keep flagging as unexplained systematics. The frame-tree formalism is developed in Paper 2, From Chaos To Common Ancestry, with the expansion-side machinery in Paper 7, From Chaos To Cosmic Expansion.
The compilation disagreements localize the same way: Union3, DES-SN5YR, and Pantheon+ differ most where environmental weighting differs, the low-redshift anchor volumes, because they are sampling slightly different patches of a real field, not erring around a fiction.
Three premises read the entire supernova sector: an environmental ratio, its local enhancement, and the frame-tree clock. The Hubble tension, the residual correlations, and the evolving-w preference are one mechanism in one dataset.
The registered kill is surgical: a Pantheon+-class analysis implementing the full frame-tree and environment correction and finding no systematic H₀ offset between cluster-environment and field-environment subsamples at more than 2σ falsifies the frame-tree contribution, and residuals statistically clean of environment after standard corrections would remove the field-sampling reading. Rubin's hundredfold statistics make the test decisive within the decade.