shadorne is right - it's about accuracy. It makes sense if you think about it. Live music is pretty loud. Music is written with live performances of real instruments in mind. Making music that was originally loud sound good at low volumes requires the response to be modified to compensate. This modification only works at low volumes, though. Increase the volume to close to realistic levels and it sounds wrong again.
The solution, in my opinion, is to build a system that is as accurate as possible at realistic volume and then use equalization to modify it when you want something other than accuracy. If you never want to listen at anything close to a realistic volume then speakers that are optimized for low volume listening are fine.
The solution, in my opinion, is to build a system that is as accurate as possible at realistic volume and then use equalization to modify it when you want something other than accuracy. If you never want to listen at anything close to a realistic volume then speakers that are optimized for low volume listening are fine.