about the comparison- i have tested the contax f2 fully open alongside the canon f1.2 50mm. the blur quality is very similar, maybe nicer on the contax- but the amount is the same.
And so it should be. Divide the focal length by the taking aperture stop. 80mm/2 = 40mm entrance pupil. 50mm/1.2 = 42mm entrance pupil. Practically the same.
That's really all there is to comparing theoretical DOF across different lenses, without changing the camera-subject distance - how big are the entrance pupils?
using the 140mm at 2.8 on the contax doesn't blur as nice or as much as a 85mm f1.2 on the canon imo. but its ok.
Again, this is entirely predictable. 85mm/1.2 = 71mm; 140mm/2.8 = 50mm. Big difference in favour of the Canon lens. If you shot with a Mamiya 200/2.8 APO (71mm pupil) at the same distance, you'd have a proper match to the 85/1.2.
its been a long time since ive used it, but i had a mamiya 80mm f1.9 that i was convinced had a really nice focus drop off- but i havent had a chance to do a side by side test with the canon.
Indeed - the Mamiya 80/1.9 (42mm entrance pupil) would be an even closer match to the Canon 50/1.2.
Now in practice, two things:
1) As Jerome has noted, aberrations, principally spherical, tweak the actual focus falloff, yielding small differences between lenses of the same entrance pupil.
2) To get the subject framing you want for the sensor format you have, you probably won't maintain the same camera-subject distance. Even using the same lens on two different formats means changing that distance to deliver the same framing. E.g. switching the 50/1.2 from a full frame DSLR to an APS-C DSLR means stepping back from the subject, and losing a lot of that shallow-DOF magic.