Remark: Scope of PlateauCurvatureSquareMatching: time-only reading and scale tensor

Theorem thm:M5-plateau-R2 is a double-plateau statement and remains true as such. It does not survive the time-only reading: with weight $w={\chi }_{\mathrm{I}}{\sigma }_{\epsilon }(t)$, or $w={\sigma }_{\epsilon }(t)$ when ${\chi }_{\mathrm{I}}\equiv 1$, and with ${\sigma }_{\epsilon }^{\prime }$, ${\sigma }_{\epsilon }^{\prime \prime }$ retained, $\mathrm{span}\{E_{w{R}^2},E_{w{\mathrm{Ric}}^2}\}\cap \mathrm{im}{\mathrm{Collapse}}_t=0$ in ${\mathcal{C}}_{\mathrm{coh}}$ and in ${\mathcal{C}}_{\mathrm{reg}}$, and $E_{w\mathrm{EH}}$ is not an image; only the ultralocal line $P_1-\frac{89}{140}{P}_2$ ($D=0$) survives (Proposition prop:to-match). Moreover the controlled scale tensor of Section sec:wpB is not in the span of the M.5 curvature targets in any reading, because its fixed $D=4$ component already lies outside (Theorem thm:wpB-dp, Corollary cor:wpB-MC); it is recorded as an unmatched entry relative to the curvature ledger, not projected.