Interface-Induced Non-Comparability Between Correct Observational Channels — A Minimal Diagnostic Stress Test

This technical note presents a minimal, abstract diagnostic stress test showing that inferential comparability between observational channels is not guaranteed, even when all channels are physically correct, fully characterized, and applied under ideal, noiseless conditions.
Through an explicit counterexample, the note demonstrates that agreement within a single observational interface does not imply agreement across interfaces, due to interface-induced non-comparability.
The result is strictly diagnostic and local in scope: it does not propose a resolution, a hierarchy of interfaces, or any general explanatory principle, but documents a structural limitation of inference induced by observational access alone.