I'll try to shed a little more light on this. For an upcoming excursion (low light work) I was weighing the 50 1.2 vs the 24-70 2.8 II on my 5D4, and decided to test both lenses in similar lighting conditions at my place. To summarize - I tried the 50 1.2 first - I really wanted to take that one because there just isn't anything quite like being able to shoot at 1.2. But, for the life of me I just could not get reliable AF going at 1.2 which surprised me - several shots dead on, then another series immediately after, complete misses. Very unreliable and gave me fits...can't afford any missed shots! Not only that, but stopping down to 2.8-4 introduced that back-focus common with this lens. Eventually I was able to come up with an AFMA setting that reduced that tendency, but the inconsistent focusing at 1.2 remained. Ultimately came to the conclusion that I don't have enough confidence in the AF to trust it (on the 5D4) and switched to the 24-70. Immediately I noticed the images were crisper and more contrasty, and this is compared to the 50 stopped down to 2.8! AF problems went away almost immediately. Didn't take long for me to settle on the 24-70, even with its considerably larger size and weight over the 50. Low-light shooting was more reliable, although having to use ISO 10,000 is far from ideal, and the look at 1.2 cannot be duplicated by many. But better an ISO 10,000 shot in focus than a missed ISO 1600 one...
Additionally, I did try the 50 on the 5DSR when I got the latter and while I found the lens to resolve a bit more than on the 5D3 at the normally soft 1.2 setting (cancelled filter works wonders), the focus issues were magnified that much more due to the high resolution. I eventually concluded that this lens is a poor fit for the 5DS cameras.
I don't know anything except a new Canon 50L can't come soon enough! And glad I'm not the only one experiencing this as I was thinking of posting something about it but thought perhaps it was just me or the camera...
:
