Remark: Single switch; preferred presentation

The preferred three-equation form keeps the physical coefficient $a_{\lambda }={\chi }_{\mathrm{I}}{\hat{a}}_{\lambda }$ (XVI-L style) and the reduced ${\bar{a}}_b={\sum }_{\lambda }{R}_{\lambda b}{a}_{\lambda }$, and collapses by ${\partial }^{\gamma }{\bar{a}}_b$ with no further ${\chi }_{\mathrm{I}}$ outside ${\bar{a}}_b$. Equivalently one may write everything in switch-free hats with a single explicit ${\chi }_{\mathrm{I}}$ in $\mathcal{G}$ and in $E_C=\partial ({\chi }_{\mathrm{I}}\hat{\bar{a}})$; both presentations are the same if and only if exactly one overall switch factor is used. Earlier drafts that wrote $a_{\lambda }={\chi }_{\mathrm{I}}{\hat{a}}_{\lambda }$ and a second ${\chi }_{\mathrm{I}}$ in the collapse $\partial ({\chi }_{\mathrm{I}}\bar{a})$ were double-counting in prose. The recorded computation uses a single switch factor: graph_general_table.txt stores $\hat{a}$ with ${\chi }_{\mathrm{I}}$ suppressed, and m4_common.collapse_column applies Leibniz once to ${\partial }^{\gamma }({\chi }_{\mathrm{I}}\hat{\bar{a}})={\partial }^{\gamma }\bar{a}$. Certificates A/B/B2 already used one switch factor; the numerical collapse matrices are unchanged by this notation alignment.