O-Series Compendium: a reinforcement of the O-Series

This is a reinforcement of the O1–O4 series, not a standalone result. It adds no new axiom and opens no new axis: it replicates the O2 core at twice the sample size, and places the human trajectories of O2 and the LLM agents of O3/O4 on a single observable, maneuverability B(t), to test one question the four papers leave implicit: are the sevendynamic regimes clusters that partition a population, or phases a trajectory traverses? The evidence says: phases.