This page does not treat any outside audit as final authority. Its purpose is narrower and more useful: separate real open problems from category errors, preserve the site's honest caveats, and point auditors to the strongest current evidence surfaces instead of letting old wording or duplicate labels do the arguing for them.
OpenTimestamps is now scoped narrowly on the site: it proves snapshot existence by time, not blanket prospectivity for every line inside a snapshot. See audit.html and methodology.json.
The public API and audit layer now tell you whether a claim is prospective, retrospective structural, supportive, pending, or open refinement. That prevents the registry count from pretending to be a single scientific score.
WIN-058 was reclassified as pending independent verification. WIN-013 and WIN-014 were reframed as non-discriminating SG nulls. PRED-SOLAR-009 remains openly unresolved and currently favors the globe direction.
The site now identifies Australia as the strongest current southern-geometry anchor rather than claiming that every global coordinate issue is already closed. See coordinates.html.
| Strawman move | Why it is incomplete | Current site correction |
|---|---|---|
| Treat every registry item as if the site still claims it is purely prospective. | The current site explicitly says the registry mixes claim classes and that prospectivity is a claim-level property, not a blanket property of every timestamped file. | Start with audit.html and claim_index.json before scoring counts. |
| Treat supportive items as if the site presents them as unique dome-only proof. | The eclipse ensemble and SG nulls are now carried with explicit competitor-model caveats rather than as decisive standalones. | See WIN-068, WIN-013, and WIN-014. |
| Quote archival AI prompt notes as if they are still the canonical read path. | The site now marks context.html as archival and points AIs to the manifest, methodology layer, and public registries first. | Use ai_manifest.json and llms.txt as the canonical AI entry surfaces. |
| Attack unresolved optics or cavity-mode gaps as if the site still hides them. | The current public pages already log the eclipse ray-trace gap and the unclosed Schumann / Tesla cavity-mode derivation as open refinements. | See model.html and methodology.json. |
| Use over-absolute kill-shot language to claim the entire site rises or falls on one non-discriminating item. | Some tests are high-discrimination and some are supportive or calibration-heavy. The current methodology page tells auditors to separate those categories before making a global verdict. | Read killshot.html together with audit.html, not by slogan alone. |