Observer Geometry I formalizes observer comparison beyond single-scalar distance. This paper separates projection, basis, and accumulation to characterize replayability, conflict visibility, and task reliability under resource limits. It distinguishes structural information loss from decoder-induced error and provides finite, checkable witnesses for core separation results, with direct relevance to distributed and provenance-aware systems.
James Ross (Thu,) studied this question.