The MOND acceleration scale tracks the Hubble rate: a registered prediction and its first confrontation

Event Density, a substrate framework, makes this reading structural and registers a_0(z) = cH(z)/(2pi) with the exponent fixed at exactly one. The exponent is the testable content: it cannot be tuned within the framework without abandoning the mechanism that produces it. The first survey able to test it has reported. MUSE-DARK III (A&A 2026) detects evolution of a_0 at ~30sigma, excluding a constant a_0 — MOND’s classical reading of its own scale, though not the only MOND (see 1.2). (The predicted local normalisation also lands within ~8%, but it rests on a coefficient nothing upstream fixes and is not offered as evidence; see 3.2 and 5.2.) The survey publishes a linear fit, not a power law (5.1). Converting it and propagating the terms the published bar omits — the intercept, the cosmology, and the reduction choice — gives alpha = 1.15 +- 0.07, putting the prediction’s alpha = 1 at 2.3sigma (5.3). The survey team’s own analysis independently describes the evolution as faster than H(z), so the direction of the tension is theirs even though the number is ours.