Boundary Information Geometry: Distinct Model Families, Preserved Failures, and a Path toward Physical Measurement

This preprint provides a source-bounded synthesis of distinct Boundary Information Geometry (BIG) mathematical and numerical model families. It distinguishes analytical constructions, model-internal and prospective numerical tests, archived-data integrity checks, external numerical holdouts, and physical measurement. Both positive and failed or inconclusive frozen outcomes are retained. The programme is not shown to have a single universal governing boundary equation. The B42 synthesis includes target-specific observability conditions and a conventional Maxwell-equivalent physical completion; no independently calibrated BIG-specific physical prediction has yet been demonstrated. An unperformed soft-interface measurement example is discussed as a possible independently assessable research question, not as a request for laboratories to validate BIG. English primary preprint, Japanese reference translation, and an English technical evidence appendix are provided. Japanese reference title: 境界情報幾何学:異なる模型群、保持された失敗、物理的測定への道筋 Version 1.0 is a new standalone preprint, not a new version of any B3–B42 source record. The English main paper is authoritative; the Japanese PDF is a reference translation of the same work, and the English appendix records technical evidence and limitations. Only these three PDFs form this deposit. Scientific scope: B42 formal overall verdict remains NONE. CA16 is an unperformed conditional measurement design. No external laboratory participation, independently calibrated BIG-specific material law, or independent full solver rerun audit is claimed. Known source/code, normalization and archive-lineage limitations remain disclosed in the paper and appendix.