Defocusing does not happen in real world shooting. The first shot of a series comes from afar back or front, but the rest of the series generally is from very close to actual focus. In testing, defocusing from both directions just adds more variables to al already difficult level of unwanted variables and permutations for us to sort through. So defocusing from a closer distance (rotate right for Canon and left for Nikon) gives more uniformity to the test shots.
If you want to test the left defocus then do it this way. Get the AFA that works best from normal testing. Then set that and take 20 shots defocusing from the right, and then 20 shots defocusing from the left. Bring them into FT and use the AFC-10 test. That will group the shots in 4 groups of 10 (or 8 groups of 5 for AFC-5) and you will be able to judge the effect of the defocus from both sides. But realize that in the field the camera naturally tries to focus from the close as it does its little dance at the end. My experience is that my test methodology is the best I have come up with as compared to the real world shooting that I do every day. Our customers seem pleased with the increased sharpness as well.