X72-02: Pure-Critical Continuous Carrier Barrier

Continuing from Round 01, where the energy route stalled at STOP-C01 (the energy/critical-scaling gap) and STOP-C02 (the vortex-stretching coercivity gap) because of its subcritical scaling, this round instead tests three purely continuous, scale-critical carriers: H^{1/2}, L^∞_tL^3_x, and the Kato/Duhamel critical fixed point. Proves that 'critical-carrier formation' is not the same as 'global critical-carrier control': the H^{1/2} route closes only for small data, and critical scale-invariance proves that the Navier–Stokes scaling cannot repair large data into small data (STOP-C03); pure energy estimates give only L^4_tL^3_x, not the L^∞_tL^3_x needed for closure (STOP-C04); and the Kato contraction method contracts globally only for small critical data (STOP-C05). The three barriers together point to one candidate transition — the next step may occur first on the observational axis rather than the underlying continuous/discrete axis — handing off to the next round's relational-geometry route.