Finite regulator

Register corrected (R81, D11, S327j). The PI ruled a correction to this row. Until then it read: statement “The 0.0331562% row is dead twice over with located cause; 0.2744490308% stands as a legal miss. • The geometry spine is separated and closed (Appendix T): "W " was two operators (Wsig preserves the KK grading, Sshell reverses it); there is no third cascade and no A4 completion; one shared radius; the transition cocycle is derived at zero continuous parameters; the two signature sheets are not glued at the F4 tier (Sylvester obstruction) but are two real structures in one complexified Albert fibre. • A four-dimensional Cartan carrier closes inside the Albert signature sector (so(5), an S 4 Cartan geometry), with metric and orientation trace-derived and triality acting as the imaginary unit; a MacDowell-Mansouri parent action for that sector follows as a candidate, explicitly not the full particle/gauge/mass action. • A reflection-positive state exists at finite regulator, together with a positive nonlinear simplicity measure - so gravity-measure existence is closed at that tier, while the pure MM/EH weight is proved not to be a positive coercive covariance.”. Nothing is overwritten (R29): the earlier row is kept in the register’s history.