Well said, Rodman99999.
I completely agree with your statement, "Testing is invariably done with relatively simple signals, generated by equipment that I'd never listen to on my system. Testing done, and empirical data collected by, listening to actual music is scoffed at. That doesn't surprise me in the least as there are multitudes in this world that are unable to hear, recognize or(perhaps) admit the differences/improvements that result from our innovations and tweaks."
I have read about Ken Stevens' audio engineering and design methods -- design, build, test, listen to music, re-design... I respect it very much! He uses testing to evaluate his equipment to determine if re-design will improve performance, but he also uses music listening to determine this. I would bet a lot of audio component engineers are designing so that test measurements will just get product out the door. I am sure this is nothing new.
I enjoyed your reply. It made me think.