The Lau Pa Sat Report is not independent evidence. It is Rich’s own carefully-constructed fictional scenario of what specific experts (Grieve, Maler, Saxe, Viotti, Smith, etc.) would say about INHERIT if they reviewed it. It was written by Rich; it reflects Rich’s pattern-matching on their public positions. It is “marking our own homework.”
Treating Lau Pa Sat as evidence-of-expert-endorsement is a category error. It can be used as a STRESS TEST (what would critical readers ask?) but not as EVIDENCE (what do critical readers conclude?).
What went wrong: v3.x handoff prompts progressively narrowed the foundation-architecture mandate from “8 candidate paradigms, decades-long horizon” down to “selective enhancement of JSON Schema foundation” on the strength of Lau Pa Sat “endorsements.” This narrowing was artefactual — it came from self-marking, not from real external review. The architectural-mandate memory (project_inherit_architecture_mandate.md) explicitly called for a 2-3 day broader paradigm evaluation that never actually ran.
How to apply:
- For decades-long architectural decisions, Lau Pa Sat weight is low. Do not let scenario projections narrow real paradigm evaluation.
- Real independent evidence requires one or more of: actual expert consultation (not imagined), published academic research from neutral authors, attested practitioner-community positions, structured peer review by uninvolved reviewers, conference proceedings, ISO/OASIS/W3C WG transcripts.
- When a prior decision was made on Lau Pa Sat “endorsement,” flag it for re-opening unless independent evidence has since arrived supporting the same conclusion.
- Use Lau Pa Sat for stress-testing proposed architectures (what would Grieve critique?) but not for justifying them (what did Grieve say?).
- The scenario-document pattern is valuable for simulated adversarial review — just don’t promote simulation to evidence.
Why: Rich designs the product and builds the mental model; scenario projections are a useful mental tool. But a 20-40 year foundation cannot rest on self-marked evidence. The re-opening triggered 17 April 2026 confirmed this.