I ran the camera through a full test set using the DeWalt target four times. A pre-repair image set is in my Nikonian's gallery except it only contains a single image at each test point whereas the actual test was three shots at each point with defocusing in between. The set consists of the leftmost, center and rightmost sensor, three shots using liveview and three shots using phase detect (normal autofocus). Prior to the repair the image set was a textbook example. All the liveview and all the center and right sensor shots were good and the left phase detect was clearly off. Obviously the most desirable result would have been to repeat the test and get the same as before except with a good left sensor. So after the repair, I did a quick version of the test the morning I received the camera which is what I showed at the top of my previous post. The next day when I had more time I went through this whole set four times and the results were just inconsistent. All of the sensors missed focus some percentage of the time. The center was notably the worst one but since the others were missing as well, it just could not be used as a valid test. Why it was so consistent before, but not now I don't know, but it could just as easily be the test as the camera. My goal is to stick to differential tests where at least one sensor is rock solid. Once one gets into testing ABSOLUTE focus it gets much more complicated. Once I switched to the USAF target the results were again consistent and all sensors were essentially the same. So, if the underlying problem is the calibration of the sensors as is the general assumption then the fact I could find a target on which all sensors were consistent and similar implies the calibration is now good. Perhaps my curiosity will compel me to try and figure out why the DeWalt test did not work this time but if it were due to the repair there would have to be something more to the left focus problem than calibration. Now maybe that is the case, nobody knows. I do recall Ming Thein commenting that he found the focus on his repaired D800 to be less reliable than his other Nikon's. In any event, with proper target my D800 focuses essentially as it should. That implies a good repair. At this point, the rest is disconcerting but potentially a red herring and not currently actionable.
I look forward to hearing results from you and Eric.