Disc galaxies show a persistent discrepancy between the observed radial acceleration and the acceleration predicted from baryons alone. A strong version of acceleration-only closure treats the observed response as a single-variable mapping of baryonic acceleration, implying that residuals at fixed baryonic acceleration should be structureless apart from measurement scatter. This paper tests that premise directly. We examine an operational structural proxy, κ = Rrim/Rd, where Rrim is an operational outer-rim scale and Rd is the disc scale length. The role of κ in this paper is strictly empirical: it is a measurable structural variable used to test whether residual organization remains after conditioning on baryonic acceleration. We do not claim a first-principles derivation of κ, nor do we estimate microscopic gain/loss rates. Using the current locked exploratory SPARC readout, we find that the residual field is not structureless at fixed baryonic acceleration. Low-, mid-, and high-κ subsets separate across shared acceleration bins, with a mean inter-layer spread of 0.104 dex. The low-κ branch drifts toward the Newtonian baseline rather than merely broadening. The claim is narrow and conservative: in this sample, acceleration alone does not fully close the disc-galaxy residual field, and an operational structural coherence proxy appears required. The decisive next gate is a genuinely gravity-blind coherence-proxy pass using imaging-only or morphology-only boundary estimators.
Mark Grant (Thu,) studied this question.