"The 50L has focus shift - as you stop down from wide open, if back focuses, but at some point (which is dependent on subject distance) the thicker DoF overcomes the back focus. So, if plotting the sharpness of the 50/1.2L based on autofocus (assuming it's AFMA'd normally, i.e. at f/1.2), I'd expect it to be sharp at f/1.2, then get progressively softer as you stop down until somewhere in the f/2.2-f/3.5 range, then get progressively sharper again, peaking at f/4 or f/5.6 then dropping again after f/11 when diffraction sets in."
So Neuro, here's the question, and I'm just spitballing. Is there a workaround for the 50 1.2L's focus shift? Have you ever AFMA'd your 50 1.2L at other apertures than f/1.2? It seems you could generate a lens adjustment lookup table for the lens based on f-stop to squeeze the most sharpness from the lens. I know you said it depends on subject distance, but the DIGIC knows what that is and could factor it in. If there is utility in this approach, the next logical step would be convincing Canon to provide the capability to enter lens adjustment values based on f-stop.