Jon, specific published figures result from what the engineers tested in the lab, and certainly aren't "lies", as some rather naive comments claim. Whether they're misleading is another matter; first, the test conditions(e.g. frequency range)may vary, and in addition there may be an implication that the very low distortion figure is of audible significance. The audibility of THD varies with the type of program material, but something on the order of 0.07% is well below audibility even on pure sine wave test tones. It's frequently said that distortion on music would have to be greater than 1% to be audible(e.g. one of the FTC regs uses a 1% distortion standard)and on complex full orchestral material the threshhold in David Clark's ABX tests was about 3%.
So, there's a bottom line which principles of scientific engineering teach us(supported by blind testing results)and which certainly isn't something as trivial as personal opinion based on casual listening. Amplifiers having flat frequency response from 20-20,000Hz with inaudibly low noise and distortion(common these days)don't create audible differences in sound, regardless of the nameplate or pricetag.
