Appendix K: Meta-lessons and epistemic audit

This appendix records what the problem-resolution program revealed about the framework and how to read it, for the next reviewer; it is an audit of process, not a new result. Of the thirteen open problems, one closed analytically, several advanced to a reproduced number, two to a forced structure, one was reframed, and one was returned as an honest negative.

This appendix records what the resolution program revealed about the framework and how to read it, for the next reviewer. It is an audit of process, not a new result.

What happened (epistemic audit)

Of the thirteen problems, one closed analytically (#3, the geometric 2π); several advanced from “mechanism only” to a reproduced number (#5, #6, #9, #10); two advanced from “qualitative” or “not derived” to a forced structure (#11 acceleration feasibility, #12 the locking dichotomy); one was reframed from a flat conflict to an open dynamical item (#2); one remains an explicit open input (#4, absolute G); and one was returned as an honest negative (#13, solar activity). No problem was closed by adjusting a definition, a value, or a gate—the audit trail is the set of scripts named in Appendix (versionhistory), each deterministic and tied to its page's slug.

Meta-lessons for the next reviewer

L1. Three honest outcomes are distinct. Mechanism-shown (a process is demonstrated), value-computed (a number is reproduced from physical parameters), and honest-negative (the framework cannot do it) are different grades and are labelled separately throughout. Conflating them is the most common way a framework over-claims.
L2. A steady cause cannot clock a periodic effect. The decisive content of #13: the solar cycle is an oscillation, an oscillation needs an intrinsic frequency, and a steady inflow has no imaginary eigenvalue to supply one. Match the time-structure of cause and effect before proposing a mechanism.
L3. Forced constants are predictions; fitted constants are accommodations. The headline a₀=cH₀/2π is a prediction because the 2π is geometric; the scorecard's grade-beside-value design (Appendix (scorecard)) exists so this distinction cannot hide.
L4. Data-limited residuals are not mechanism failures. #6 (full χ²), #7/#8 (joint fit), #9 (η–δ degeneracy), and #10 (absolute value) are bounded by missing data, not by the framework, and are marked as such rather than presented as successes or failures.
L5. Out-of-scope is not unsolved. Cosmic origins (nucleosynthesis, the acoustic peaks, structure-formation history) are excluded by design; stating that plainly is more honest than manufacturing a closure the present-facts standard cannot support.
L6. Reproduce from the slug. Every quantitative claim ties to a named script under repro/cosmology/{slug}/ with the same slug as its page; the provenance ledger (Appendix (provenance)) marks each number geometry / simulation / measurement. A reviewer should start from the scorecard (Appendix (scorecard)) and follow the grade to its script.