The same pro system in a home theater is not likely to experience compression, even with very high power levels occurring from time to time, but in a rock concert it may well see compression.

Either garman or I am confused about the results. If the voice coil heats up and compression comes into play, then both volume (maximum) and dynamic range are affected. Whether or not the effect on dynamic range is linear, I don't know. If maximum spl is reduced by 3dB, then spl should be reduced at every non-zero power level input by some value less than 3dB but greater than zero. This is what is meant by "dynamic range," isn't it? If the effect turns out to be linear, then there is no real effect on the music, except that it is not as loud.