Ok - did the Thom Hogan tests for the left focus issue. I shot at targets on the wall from about 10 feet. 105VR macro, f2.8, mounted on tripod, orientation of lens as near perpendicular to the wall as I could get it. Camera dead level according to tilt sensors. Mirror delay to reduce vibration. Lens AF corrections off. VR off. Lighting as bright as I could get it. 1/250th at ISO 400. Results were interesting:
1)Doing such a test handheld has no chance of meaningful results.
2)Af is much more precise in Live view than phase detect. sharper and more repeatable - should this be the case?
3) no discernable left focus issue when the camera is mounted on a tripod. Left right or centre sensors gave much the same focus right across the frame but all three marginally less sharp than using the contrast detect system in Live View.
4) some "hysteresis" in the mechanism. when you focus down from infinity or up from min focus distance the results are slightly different. focusing down from infinity was better. This could be due to the lens mechanism, not necessarily the camera.
5) don't use VR when mounted on a tripod....my first set of tests had to be trashed because of this oversight.
Much as there are some doubts about the true history of my "new" camera, it seems to be working OK. the slight issue with Phase detect AF being slightly off compared to Live view can be fixed with a lens correction since it seems consistently off across the whole field of view. I previously had a small lens correction in my D300 for the same lens.