You don't need to take so many pictures to find the sharpest AF setting (AutoFocus Adjustment or AFA in Michael Tapes terminology). Michael Tapes mentions a 5 step test shot session to get an overview, then a 1 step refinement at a smaller AFA range.
Depending on whether you have LensAlign or not, the number of required test shots differ. There are two examples explaining this in the guide (chapter Analysis Techniques, paragraph AFA Autofocus Adjustment).
I added the number of shots needed in the text (=>), based upon 4 shots per AFA, to get a better average.
analysis of the entire adjustment range => 60 shots when using flat target • start with taking shots at AFA (-20 -15 -10 -5 0 +5 +10 +15 +20) => 36 shots • supposing the Autofocus Adjustment Analysis says -10 to -5 are sharpest, proceed with extra test shots at -10 -9 -8 -7 -6 -5 (=>24 shots), and reprocess to find the best AFA value.
analysis of a part of the adjustment range => 45 shots when using LensAlign • take a test shot at 0 AFA => 1 shot • view the LensAlign® ruler picture to determine back or front focus • in case of back focus, take test shots at AFA (-20 -15 -10 -5 0) => 20 shots • in case of front focus, take test shots at AFA (+20 +15 +10 +5 0) • narrow down with the help of the Autofocus Adjustment Analysis • take a series of shots in 1 step increments and reprocess => 24 shots
If the graph is pretty flat, you can narrow down the analysis even more by removing the less sharpest AFA's in the FocusTune grid, as I did with the AFA step 1 demo file (removed -10, -4 and -3), and then reprocessed them once again.
The bump in the first graph, between -10 -5, suggests there are better values. The analysis tool does not make a prediction though, it is merely a way of connecting the dots with a smooth line*, a feature of the graphical tool being used. It doesn't mean that you will find a better AFA exactly between -10 and -5. You will only figure out by digging deeper by means of 1 step shots.
BTW, Michael Tapes announces a new guide in his posts, the current is a draft version as you may have noticed. Maybe this week?
To conclude, I found FocusTune easy and straightforward to use. I followed the recommendations as far as possible, thereby using LensAlign. Took my shots indoor, as the sunlight may fluctuate during a test session. The more stable the light on your target, the more trustworthy the result is.
* also straight lines from average to average produce a fake prediction. The smooth line is nicer, even if the line doesn't really matter. The line makes the graph easier to interpret.