Upload & Sell: Off
I have never seen anyone testing for decentering on such a short distance, and I sincerely doubt you can be sure anything you observe is not because of the sensor and the test target not being perfectly parallel. Doesn't one usually test such things at infinity, for example using a horizon going from the top left to the bottom right or from the bottom left to the top right corner?
As for the lens: I have had a couple of copies and sent them all back. Some were clearly decentered. The two copies that were not clearly decentered exhibited what looked like extreme field curvature at large distances and with short FL. This meant subjects at infinity placed at the short borders could never be in focus at all for 16mm and f8. It's far from horrible, but also far from sharp. I am still not sure whether these copies performed as they should, as in Kurt Mungers review I see better performance. So it may be a problem with build quality and quality control. I really wanted reasonable WA performance at large distances, which is why I gave up on the 16-80, planning to try the 16-50 (which was just rumored at that time).
Most reviews stating the 16-80 has really good corner performance use imatest, I think, and at short distances the lens can be quite outstanding. FC disappears, and the corners should become really good.
OT: BTW, I have a similar problem with the 16-50 now, of which I have had two copies with truly outstanding corner performance on one side of the image. Unfortunately, the other side was rather horrible. I had decentered copies of that lens as well, btw. The construction may feel much more solid than that of the 16-80, but I have yet to find a single copy that is actually assembled properly.