We present an audit-grade formulation of the Quantum Information Copy-Time (QICT) program as a micro–macro closure framework and as a quantitative pipeline for falsifiable predictions. The core observable is the operational copy time τcopy(ℓ;ε, δ⋆): the minimal time required for a calibrated local bias in a sender region to become statistically distinguishable in a receiver region at separation ℓ, under explicit signal-to-noise accuracy ε and disturbance budget δ⋆. Under transparent hypotheses—locality, sector ergodicity, and the existence of a quantitative diffusive hydrodynamic window—we derive a one-way lower bound τcopy ≳ ℓ2/D with a strict feasibility correction controlled by the inversion of the diffusive tail. We show how a measurable Spectral Diffusion Criterion (SDC) in the hydrodynamic sector converts microscopic unitary dynamics into an auditable transport closure. We connect this closure to two predictive targets: (i) an inertial spectral mass diagnostic defined from long-wavelength spectral flow, and (ii) a reproducible Higgs-portal dark-matter corridor in the scalarsinglet model, where the QICT calibration acts as a restrictive prior on the effective portal region. A complete reproduction package (code, data products, and figures) is provided; we emphasize which statements are definitions, which are assumptions, and which are falsifiable predictions.