[LittleTester]

Well in the most common family of interpretations "collapse" isn't an actual physical process, just Bayesian updating. So you wouldn't expect to find physical evidence of it in that sense.

It's true that from the perspective of an external superobserver the quantum state evolves to contain terms for each observer observation state. However since all interference observables turn out to be non-physical for macroscopic systems we get a superselection rule and so the probabilities for different macrostates are classical probabilities and thus reflect simple ignorance of the observer's post measurement state.

There's very little motivation for reading the quantum state "ontically" in the way you are doing.