You do what you can and where you can.
Today on the DAC lv you can have it for cheap, and same goes even more for small amplifiers (for hedaphones) or Preamplifiers (116~122 dB SINAD), you lose some 20 dB from that with power amplifiers (and rather good one's 100 dB SINAD) and additional 7~8 with great (96 dB with ±2 dB 60-20) speakers. With headphones you can get higher 107~108 dB ±2 dB 20-20.
To answer your questions because we can and are spoiled brats who like to bring in loads and loads of (controled) distortion (or even less controlled one even not harmonic [beats me if people love it and some do]) into the reproduction chain.
This is a good argument and I try to put it in numbers.
I was aiming for CD quality, that means 16bit 44.1Khz Stereo sound, with SNR around 96dB or more. I knew about THD+N, so i picked integrated amp with better value 0.005%.
I ended up with better equipment on audio ouput side I bargained for - 192Khz, 24bit with roughly 98dB SNR, but because the source was Windows, the THD+N went to 0.1%. But after some tweaking and testing i was able to get measurable results where THD+N got to 0.005%.
Doing some blind tests i cannot tell the difference between Shared and Exclusive mode anymore, but Windows does more bad things to audio signals than just bad VolumeLimiting or sloppy resampling.
Looking at the distortion graph for DBR-62 speakers I just assume I should not be able to hear any difference just based on distortion most of time.
THD is most of time above 0.1%, so the difference I hear... its either not THD, some other artifacts, placebo or is it possible that listening to those speakers on lower volume than 86/96dBSPL also decreases THD they produce?