Agree with everything above by m-fine and I was going to post the same things as Danny about the low SPL instants in between high SPL instants.
Here's my understanding of that bit, correct me if (where) I'm wrong!
Snock, to answer your thought experiment (without having heard any GR speaker yet), imagine a speaker with good dynamic range and one with poor dynamic range, both set to the same SPL level, both comfortably within their limits, both happy, at medium volume.
Now imagine a Double Bass or Cello note, one string is vibrating, centered at say 100Hz. The sound on the source material is theoretically going from quiet to loud, 100 times per second. The speaker with bad dynamics isn't actually going very quiet though, it's staying louder during the quiet bits, adding it's own sounds that shouldn't be there.
Both speakers will hit the same peak SPL, but the one with great dynamics willl hit a lower minimum
SPL during that note, if it were possible to measure that (you would need a pretty fast DB meter!
Great dynamic range sounds like silence between the individual vibrations of the string. It sounds like better resolution, more detail etc, etc, but really, it's just better dynamic range.
Capacitors have the same effect as Danny said. It's about starting and stopping the music quickly, and only reproducing what is in the recording, not all the other sounds that all systems introduce on their own.
Does that make sense?
Out of curiousity, is it possible to measure the SPL range of a note at say 100HZ (or 10'000HZ) and see in a graph what the SPL level does 100 times (or ten thousand times) a second? Does equipment to do that exist?
Could that be the area of sound that we can hear but not measure? I'm thinking of cable debates and things here.