I'm not sure that you have to worry too much about the film thickness building up on the takeup spool changing that much. I have had a number of folders with auto functions on the takeup.... Balda Baldax (2 of them), PerkeoII (4 of these), Mamiya Six folder Auto (3 of these). Of these, the Baldas were the only ones that did not break down. The Perkeo's were about 50/50, and all of the Six Auto Mamiya were total failures after a while. However, while they were working I can tell you that there was never anything very sophisticated to take into consideration build up of film on the take up. The fact that the 220 film is thinner means less buildup as you go along until you have almost the same amount of film on the takeup by the end.
My point here is that I think you can do quite well counting the number and degree of turn, and then giving it a nudge for good luck. A bit of excess spacing can't hurt. And you do have to cover that ruby window, so counting turns is your only alternative. Also, the unique film pressure plate in the Mamiya Six folder would certainly take care of any differences in film thickness. The plate slides in after the film is loaded and is held firmly against the film whatever the thickness. This and the moving film plane focus is what gives the Mamiya folders a HUGE advantage in film flatness over all folders I have used. I would love to figure out how to adapt that system to a 6X9 folder.
Also, On both the Perkeos and the Mamiyas that failed to wind film with a locking counter, I have found it easy to strip out the counting mechanism, revert to the ruby window for spacing and still keep the double exposure prevention and the automatic shutter cocking. The Mamiya is extremely nice once you get rid of the stuff that breaks down, because it can be manually cocked for intentional double exposure. Using the ruby window has always given better film spacing on 120 film than any of the other auto mechanisms other than the Balda.
