Since I use a 4x5 camera on a compound microscope (microphotography), I would say no rule of thumb applies, certainly diffraction has not caused me to use 35mm. And the trouble with your hypothesis is that you want to create an artificial playing field by ignoring granularity and view camera movements. You would also want to think about numerical aperture as well. Things are never equal. Use the tools that give you the result you want in the way you want. Whatever format you use, you will always be making compromises.