Expect subtle differences between any two shots regardless of the AF system used. If you do not see a dramatic difference, consistently and repeatedly, you have no problem. When it is there, it is not subtle and left most FP using Phase Detect AF will be seen without blowing up to 100%.
Many of the sets posted where the owner claimed it was bad showed nothing conclusive and took extreme magnification to see. Some coming from DX were not prepared for seeing general softness at the corners or sides on otherwise very well regarded lenses. A 24-70 or 85 1.4 or other well regarded lens is not as sharp on the sides compared to the center. That was not seen on a DX camera since projection onto the sensor is so much concentrated in the center of the Fx lens's image circle. It is interesting to see the galleries of people who were sending one camera after another back, seldom used outer FPs in real usage. They likely never tested their other cameras and did not see that all of them would have been less than perfect off to either side. I did not see any problem in actual photos in normal use so never tested it but when I did when I suspected a problem with one wide prime later one, the results simply confirmed what I had seen in all the images, it was fine. I only suggest bothering with testing thoroughly if one understands the systems involved and can interpret the results of tests realistically considering the weaknesses inherent in home testing with too many uncontrolled variables. Shoot with it, if you see a consistent problem that is not explained by user influences, where the images fail subjective tests at normal viewing sizes and magnification, then explore why with well conducted tests. If you do not see a problem at normal viewing sizes you do not have a problem. The ability to see some detail at 400% x is causing more claims of "softness" than the feature benefits. Stan St Petersburg Russia