Finite-Window Local-to-Clean Transfer and Anti-Phantom Detection for Sharp Navier-Stokes Packages

We prove a fixed finite-window structural theorem for sharp localized Navier-Stokes packages, formulated as both a local-to-clean detection theorem and an anti-phantom principle. The result addresses whether a defect visible in the baseline quotient geometry can disappear after pressure-tail enrichment, residual transfer, quotienting, and clean-to-local detector comparison. Under synchronized representatives, baseline-to-tail visibility, component comparison, residual-ledger closure, detector comparison, chart visibility, and a clean quotient gap, any baseline-visible defect is either detected by the localized detector or charged to an explicit quotient-residual ledger. Quantitatively, M_Lambda^loc(D-zeta_*) >= c_{Lambda,0} Dist_{loc,intg,0}(D,Gamma^{intg}*{Lambda,adm}) - E^{quot}*{Lambda,0}(D). The proof assembles three modules: pressure-tail visibility, componentwise residual-ledger closure, and detector comparison. The anti-phantom interpretation is that a baseline-visible defect cannot be simultaneously detector-silent and residual-cheap. We also record provenance for the imported quotient interface, finite-dimensional pressure-tail models, explicit matrix realizations of the structural inputs, NS-generated coordinate realizability, compactness criteria for clean pressure images, and reduced pressure/tax kernel-free criteria.