CO-EMERGENCE AS CONSTITUTIVE DYNAMICS — verification
=====================================================
Change: folded three-face force C(Ω) into phi_pool's per-cell amplitude update
        (step_wave), so co-emergence is part of the tick, not an after-check.

C(Ω) derivation (phi_pool_coemergent.c, coemergence_force):
    u      = A / √φ                     coherence variable (A at FIRE ⇒ u=1)
    r_sq   = u² − (u+1)                 square vs translate face; 0 at φ
    r_inv  = u − (1 + 1/u)              T fixed-point face; 0 at φ
    C      = −φ⁻¹·∇½(r_sq²+r_inv²)/√φ   φ-gain, bounded ±0.5, floor u≥0.25

RESULTS (GRID=32, genome 0x5625BA88, settle 120, deterministic, 2 runs each):
    baseline    : kur_R = 0.1020   state = PLUCK     (uncoherent)
    co-emergent : kur_R = 0.4813   state = SUSTAIN   (coherence forming)
    => ~4.7× phase coherence; field advances one lock state; genome preserved.

Field liveness: 4096/4096 slots live (no collapse), full field intact.

KEY FINDING answering "after-the-fact vs part and parcel":
  Co-emergence is DYNAMICAL, not descriptive. It belongs in the evolution law
  (glyph FIX, term C(Ω)), not in DERIVE as a comment. The fixed point lives in
  PHASE θ (S¹), not in the 𝓛ₙ amplitude projection — so C acts through the
  amp→phase coupling, which is why kur_R (phase coherence) is the metric that
  moves, and why measuring amplitude three-face residual showed no change.
  The substrate now self-organises toward φ every tick rather than being
  audited against it after settling.
