Please stop uncritically repeating this threshold figure. 10% distortion is -20dB. That's very high and falls well into the main dynamic range of music.
Sure, there's a lot that goes into perception, not the least of which is that many people don't even know what distortion sounds like and have no personal reference for clean sound. It's also no good to have unrealistic distortion expectations for your equipment. But that's no reason to repeat random things like "0.1% is inaudible", "1% is inaudible".
https://secure.aes.org/forum/pubs/journal/?elib=5147
View attachment 107362
I agree. Rythmik is good, but they provide scant distortion figures and other measurements. Measured group delay for their subs and others according to Data-Bass is roughly the same. This also goes for the CEA 2010 results: good, but there's no evidence of a wide gap between servo and non-servo designs. My impression is that the main effect of the servo system is a more even frequency response across a wider range of levels, and that at least some of the improvements it offers are compensations for driver deficiencies. It would be very informative to compare recent Klippel-assisted sub driver designs with the Rythmik drivers, and we already have the latter because of
@hardisj's F12 review.