JMan, thanks for presenting this, I've enjoyed it.
Kudos to your model, by the way; I use subjects such as trees, and sometimes I can't convince even them to stay in the same place.
I'm still thinking over how the format differences cause the different catch-light rendering between formats. That there's a relationship between the absolute aperture (not f-stop, but 50/2.8, 35/2, and 25/1.4, or roughly 18mm) and the format size, which affects rendering aside from brightness and depth of field, is an idea I hadn't seen in the format-equivalence discussions I've seen so far. I still haven't wrapped my head around it, but I'm grateful for this prompt for thought. So, again, thanks.