>Basically, after running through this program on my 24-70 @ >70mm @ 3.5M approx 10 times I have no level of confidence in >the results as they are completely not repeatable.
Did you look at the results? I had that experience with one lens (200-400) that resulted in a very flat curve, i.e. going from -10 to +10 made very little difference. And indeed on the images it was very little difference. This flat curve means that, statistically, the results are a bit chaotic as small variations in each test run cause the peak (of a near flat line) to vary widely.
Other lenses had a very sharp peak, a clear maximum where +/- 5 made huge differences in focus.
I found I learned a lot by these curves, more so than the actual adjustment it called out.
Your experience may be quite different however, just offering one result.