I was refering to testing tuners using full revolutions instead of micro adjustments ex: 3-5 hash marks at a time.As for tuners... I'm not getting into that here except to correct your statement that we didn't follow the manufacturers recommendations. We did that, and more. The thing we did that's not in the instructions, is to see if the instructions result in the same 'best' tuner settings when you repeat the test. In fact the results do not repeat, meaning each time you do the test you find a different setting. This conclusion has been reached by everyone who's tested tuners with statistically significant sample sizes, and checked for repeatability.
Anyways it seems the "repeatability" of tuner settings is a different and seperate test than a test to see if tuners can change/improve accuracy. For example, perhaps the tuner's spring/tension system/design is not perfectly repeatable via engraved hash marks, but it still is able to tune the precision. Maybe last time it was 7 hash marks from zero but hafter fiddling up and down back and forth especially considering you tested several full revolutions at a time unlike any manufacturer's instructions, now it's best at 9 hash marks instead of 7. The setting maybe wasn't repeatable but it still tuned the precision. I don't know if you or anyone else tested that.