The mechanism you offered based on your experience seemed to fit and I expanded on that thought. I don't really see the need for your aggressive response, to be honest.
No, no, I am not trying to offer "this mechanism," from our portrait camera as a possible explanation for the Medalist problem. (I did not actually explain the mechanism behind our portrait camera problem.) I am offering a TEST METHOD to check for film flatness. I only brought up the portrait camera issue as demonstration about the effectiveness of the "flatness check" method.
Ps, until you revealed that you had switched the topic from (my) portrait camera back to the Medalist, it looks exactly like you were telling me that I didn't know what size film I'd been working with for years. I reread things carefully and decided that things had to be made clear. I get now what happened; I don't think any apologies are called for.
FWIW I deliberately avoided describing the test method as it might seem too unbelievable (and/or complicated) to the casual reader here. So I prefer that someone be interested enough to actually try it out, in which case they'll come back and say "OMG, that's almost unbelievable how well it works!"
The test method is somewhat akin to something those "paintless dent removal" people do when repairing hail damage to a car. They need a way to judge the effectiveness of their efforts. So they place a small light box with a dark grid pattern on the other side of the dent. Then while looking at the dent they see a distorted reflection of the grid, and minimizing this distortion is their guide while gradually removing the dent.
Here's a conceptual description of the film test method. One attaches a grid pattern to the front of a light box. Then set up the camera as if photographing the grid. (If the shutter is opened an image of the grid is projected onto the film, right?) Now for the tricky part(s): put a small mirror in front of the camera lens at a 45 deg angle, and turn the camera so as to again be photographing the illuminated grid. But... instead of a normal mirror, use a so called beamsplitter where, for example, it might reflect half the light and let the other half go through. (Window glass behave as a beamsplitter except that it reflects only about 10% whilst about 90% goes through.
To evaluate: we know that an image of the lit grid is being projected onto the film, via the beamsplitter, right? By looking into the front of the lens, through the beamsplitter, one can inspect the image being projected onto the film! If there are bulges or other distortions in the film surface they will show up as distortions in the grid. Does it work? Yep, works great!