Ken,
I don't think anyone is advocating playing a system at a steady 110 db, but the ability to reach that kind of dynamic, if only for a millisecond makes quite a difference in presenting a realistic sound picture. back in the late 80s some did an experiment (I believe it was J. Peter Moncreiff of IAR) that used an oscilloscope to look at the microphone feed of the click made by a pair of scissors opening and closing. He recorded this and then tried to reproduce the same sound through a spare of speakers, at the same SPL as the original.When he looked at the output of the amp on a scope he saw that the waveform was clipped as compared to the input signal. He kept substituting bigger and bigger power amps until he found one that didn't clip the output (again with the same spl.) It wasn't until he got up to nearly 1000 (yes 1K) watts that the signal out of the amp was a faithful reproduction of the input signal. Now granted, it only needed that amount of power for a mere fraction of a second, but without that kind of headroom you could not reproduce the full fidelity of the signal. And while there are not too many scissor clips in most of our music collections, I dare say that they do contain a lot of cymbal crashes, which have a very similar transient profile. The bottom line is that the advantage of a big amp is not that it can drive your system to higher volume levels, but rather it can make your system sound better when it is playing at lower volume.