Experience with lots of pro and hi fi systems over the years says low level sound quality is mostly related to higher gain audio quality- as simple as that. Despite the obvious Fletcher Munson ear sensitivity issues, that would point some towards a loudness control, most of these loudness controls are not calibrated to the efficiency/SPL of the loudspeaker at a given volume- so are just a random more bass and treble control (which when used always sounds better to most listeners). But it is not more accurate. Fletcher Munson curves are directly driven by SPL so must be calibrated to the system and its output to be accurate.
In pro mastering, where a lot of mastering is done at low level and high level both, there is no such thing as a different EQ curve for loud vs soft. I’ve never met anyone in 30 years who thought they needed it or asked me if I knew of anything that did that. If the speakers don’t sound good at low level, they probably didn’t sound good at high level either. Experienced listeners accept there will be less low and high end at low level, but they expect the midrange (the translating part of the sound) to be consistent between low and high level. Engineers and customers both expect the mix or master to sound good at ANY level.
NOW if I am only listening at low level, that could be a different scenario. I might choose a completely system all together. But it wouldn’t be for improved accuracy- it would be for better enjoyment!
I think this whole issue is not worth any jump to DSP as a method to cope with level. The move to DSP can completely change the perception of the entire rig as it affects everything, not just low level loudness compensation.
Brad