The Audit Asymmetry Problem: Capacity conditions for substrate-independent audit

The first note in this series identified a class of evaluation failure — category-level failure — and proposed three structural conditions for substrate-independent audit. It closed by flagging that its information-theoretic argument rested on an analogy to channel capacity rather than a formal result. The present note takes up that open question. It argues that the conditions under which substrate-independent audit ceases to be informative can be specified along three dimensions of an audit relation's capacity — bandwidth, semantic reach, and self-reference depth — held relative to the requirements that the audit task induces from the auditee. When the auditor's capacity falls below the task-relative requirement on at least one dimension, and no compensating institutional condition restores the missing dimension, the audit cycle's judgments cease to track the task-relevant state of the auditee. The note develops each dimension through a formal source whose role is to mark the shape of a constraint, not to derive it: Shannon (1948) for bandwidth, Rice (1953) for semantic reach, Gödel (1931) for self-reference depth. It reads the 2008 Gaussian-copula collapse as a forensic instance of capacity asymmetry in five stages, and the LLM-as-judge architecture of current frontier evaluation as a contemporary instance. The first note's three failure conditions are reframed as duals of the three capacity dimensions. The note concludes with the asymmetric substitutability of the three externality conditions: provenance independence and external anchoring can be approximated by procedural repair; speed parity cannot.