What unsupported assumptions have you mentioned about Type 1a supernovae, for instance?

I don't know why you linked to Merryman's comment, either. His objection to photons being stretched along with space makes no sense--you don't get ordinary translational redshift either if photons aren't wavelike in that regime. Doppler radar works, so....

(Also, you didn't answer the objection in my other comment about the flaw in your interpretation of the raisin model with regards to density, if space is expanding rather than having galaxies translating away. I mean, you really have the same problem in both cases--the real issue is where you decide to say that the galaxy "is" once you see it. In the translation case, "Where the galaxy was then!" seems the obviously right answer, whereas in the space-stretching case, "With the current stretched geometry!" seems a more reasonable choice.)

One could make an interesting point about angular extent vs redshift vs apparent brightness in the standard model in contrast to, say, a tired-light model. But you need to address the issue in a more technically detailed way, first. (Edit: I am not endorsing the tired light model, at least not as the primary mechanism of redshift. It doesn't explain supernova duration broadening that is consistent with relativity. But there are things that the standard model is not very good at predicting.)