My results are also too inconsistent to post, but since I am using the Mac Beta I think it fair to wait until a final release before drawing strong conclusions. In the meantime I have ditched my slightly Heath Robinson Multi Point chart and made a new one by A3 printing the chart on thick matte paper multiple times and trimming it, then carefully mounting it dead flat, several abutting sheets, with double-sided tape onto a 1.6 x 0.8 metre sheet of very thick, flat board - so I know I will not be able to blame my target!
I found that with the single point target, as long as I got the settings and setup correct, I got almost identical results as to those achieved by using a Spyder, with all my lenses. The multi point test results are less consistent but if I use High Focus Consistency and many shots of data point, they are somewhat better. I even got a few that were all green, but always with AFFT values for the centre point that differed from those achieved using the single target and were not useful in real-world shooting. When I ran the test again, it would differ. This is a bummer because running the multipoint test with, say ten AFFT values and a high consistency requirement is very time-consuming and uses a LOT of actuations.
But as I say, it is beta, so I won't hang them yet.
One thing I must mention is that you are using the wrong lighting and this makes your results useless I am afraid. Focus of red spectrum light is further back than mid-band. This is one of the reasons Leica Monochrom users have problems when shooting with a red filter on. It's only of use if you intend to shoot under red light! So I always use scrimmed daylight, and this makes a significant difference: such that, for the single point fine tune test, I absolutely trust my results and they are very consistent, though they do vary depending on whether my testing room is cold or warm...
