Ravi -
I believe you are correct when referring to "dynamic headroom" as the ability for the amp to product large increases in volume for relatively short bursts.
The reading that I have done suggests that the "dynamics" of dynamic headroom are different for home theater vs. music (usual argument is that home theater is the more demanding as it requires longer, sustained peaks).
As many reviewers point out, separate components are typically designed to a higher standard. That's why a 60W "separate" amp will outperform a 60W AV receiver. I suspect a large part of the increased performance of separates has to do with its ability to generate dynamic headroom without distortion.
So, to your question about distortion at sonic peaks in program source material: At the same volume level, I suspect that the difference between two AV receivers (one at 75W, one at 100W, for example) would be almost exclusivley explained by the amps' dynamic headroom.
Few manufacturers dare report dynamic headroom, and I have never seen a reviewer attempt a test to measure it (if indeed it can be "measured"). Yamaha is reporting it (see for example, http://www.yamaha.com/cgi-win/webcgi.exe/Specs/?gAVR00010RX-V730), but I have no basis to know if the spec they report is good or not (they claim 1.2dB of dynamic headroom at 8 ohms). This doesn't seem like much, but I suspect, given Yamaha's good reputation, many AV receivers would fail to achieve the same dB increase.
The other interesting number would be how long an amp could sustain its dynamic headroom spec.
