I measured 8 degree difference on an old Lucas K2F mag due to bad armature alignment... Many magneto servicing places address the electrical issues but not the mechanical issues
This is a very important point, and you clearly understand the issue.
I didn't measure the slope of the ramps on the Bosch's cam, but on a Lucas
K2F it is ~2.4-deg. of engine rotation per 0.002" of lift (1.2-deg. of cam rotation). That means that even if the cam is made perfectly, and you set the points to open on one cylinder precisely where you want them to, but if the cam is held in the housing such that its centerline is offset from the centerline of the armature by only 0.002", the other cylinder will fire 2.4-deg. too soon or too late.
Magnetos are electro-mechanical devices, so even if the electrics are in perfect condition, that only takes care of the 'electro-' half of the problem. As a preview of a future installment, one of the upcoming tests will be of the restored magneto run on my modified distributor tester at 1250 rpm (2500 rpm engine) to determine precisely where both cylinders fire during actual operation, and to see how much variation there might be during operation (due to, say, sloppy tolerances).