The radial acceleration relation (RAR) is organized by an acceleration scale near 1. 2×10−10 m s−2. We test a conditional construction in which a declared present-horizon protocol reads one massless thermal budget on two matched supports. Its exact support-rescaling identity gives alead = cH0/ (2πΩΛ3/8) = 1. 2005×10−10 m s−2 for Planck 2018 inputs. The exponent 3/8 = (3/2) (1/4) combines the flat-FLRW support-volume ratio with the three-dimensional thermal budget. For the data-facing direct branch we additionally declare aSC = alead and unit interface coefficient ν = 1, so that aRAR = alead; these normalizations are fixed before, not fitted to, the galaxy comparison. A separately declared minimal closure then gives gobs2 = gbar2 + aRARgbar (1 + √ (3/2) ·√y) e−y/2, with y = gbar/aRAR. On the SPARC Q = 1 sample (93 galaxies, 2060 radial points, fixed catalogue mass-to-light ratios), an effective-scatter likelihood with galaxy-block uncertainties yields aRAR = (1. 213 ± 0. 084) ×10−10 m s−2. Conditional on Cg = 1, this corresponds to ν̂ = 1. 011 ± 0. 071, consistent with the declared direct value at 0. 15σ. With aRAR = alead fixed, the separate readout-amplitude coupling is Cg = 0. 996 ± 0. 064. The envelope coefficients are consistent with but weakly constrained by the data; fitted alternatives are statistically unresolved under galaxy-clustered comparisons. The construction is not an unconditional derivation from partial observation: the horizon, thermal state, support pairing, anchor and interface normalizations, channel list, and profile branch are explicit gates. The SPARC analysis is a retrospective fixed-catalogue consistency test, not an independent empirical certificate. The selected bundle is falsifiable through its gates, including a positive-semidefinite one-sided response, a gate-conditional ℓ ∈ 0, 2 sky pattern, and an exponential high-acceleration tail on the source-rate branch. This record contains the Letter and its Supplemental Material (exact support trace, complete gate ledger, envelope derivation, likelihood audit, robustness tests, and reproduction instructions) as a single merged PDF, together with the arXiv LaTeX source package. Theory companion paper: doi: 10. 5281/zenodo. 21423426. Preprint version v0. 8; submitted to Physical Review D (Letters). v0. 8 (2026-07-25). Changes below are stated relative to the previously published v0. 6. 2 record. Title. "prediction" is replaced by "anchor" and "SPARC test" by "SPARC consistency test", matching the epistemic scope stated in the manuscript. Notation. The single symbol a0 is replaced by an explicit three-role dictionary: the support anchor alead, the optional stochastic anchor aSC, and the galaxy-facing scale aRAR, with the interface map aRAR = ν aSC exposed as a declared gate rather than an identification. New material. An "Empirical certificate status" subsection states that the SPARC calculation is a retrospective fixed-catalogue consistency audit and not an independent empirical certificate, and audits two prospective branches (the public galaxy-flaring branch is not executable at the required joint error level; the WALLABY pilot supplies independent galaxies and usable points but lacks the stellar-mass, molecular-gas, distance-posterior and beam-covariance packages). An optional Ornstein-Uhlenbeck anchor branch is registered separately and differs from the direct branch by under one percent. An intermediate-redshift continuation check against MUSE-DARK rules out treating the present gate bundle as an epoch-local law. A "Producer status and non-identification" subsection states that the selected exponential envelope is one member of a conditional profile family rather than the unique output of the support calculation. The Supplemental Material adds a "Physical rationale for the selected branch" subsection covering the no-wall budget, present-epoch clock, matched support ratio, unit reference normalizations, leading-mode closure, isotropic coupling baseline, and the source-rate weak-monitoring branch. The introduction now states the branch-selection economy principle compactly (no new acceleration scale, unit reference normalizations, minimal positive analytic closure, all declared before the data comparison). All alternative walls, clocks, pairings, normalizations, channels, and exposure laws remain explicit registered branches. Reporting. The conditional calibration ν̂ = 1. 011 ± 0. 071 and the two-parameter Wald statistic W = 0. 87 (p = 0. 65) are now stated explicitly; the equivalent 0. 46σ was already reported in v0. 6. 2. Files. The Letter and its Supplemental Material, previously two separate PDFs, are now one document, and the arXiv LaTeX source package is archived with the record. Unchanged. No equation, dataset, likelihood, or empirical verdict changed. Statistics are unchanged from v0. 6. 2: aRAR = (1. 213 ± 0. 084) ×10−10 m s−2 on SPARC Q = 1 (93 galaxies, galaxy-clustered errors), 0. 15σ from the predeclared anchor.
Yi (Sat,) studied this question.