BTW, the "off the shelf" modules you talk about were listened to and tweaked over a couple year period by various people in the high end community. Check out Mark Shifter’s comments on this (he was one of the main people). They made it sound the way it does on purpose.....I think they did a very good job. The measurements were secondary to the listening tests. And that is just the way it is. Believe it, or not. he he...
Well that is interesting, considering that these are essentially variants on the GAn Systems Evaluation boards that Skip Taylor designed for GaN systems, but with some of the flaws fixed and now sold under his company name. Considering those came out about March 2020 and the LSA GaN was done in July 2020, that "couple years" comment is quite suspect. It does not "sounds" any particular way. The distortion while not state of the art (except for IMD) is below audible levels, so that pretty much puts us at how the output impedance goes up at higher frequencies resulting in speaker dependent objective and subjective differences.
Mark Shifter also said this,
"We believe that GAN FET technology is the holy grail of switching amp design"
... which has about as much relevance to audio and amplifiers as saying "I like blue".
Since you don't appear to have anything relevant or concrete to add to this conversation @ricevs, I will say good bye.