sorry, this one's really long (again), but i think it clears things up for both sides of our little debate.
ok, i got to thinking about "gain" and how that could play a role in blowing your speaker. here's the deal. increasing the gain on an amp increases the volume of the entire dynamic range until the amp starts to clip, at which point the peaks begin to be cut off. if you keep cranking the gain, your peaks do not change volume, but the lowest part of the dynamic range becomes louder, while the middle part starts to become the "peaks" in addition to the original peaks. Think of the signal going into the gain stage as a piece of paper hitting the ceiling of your room. as you push it harder and harder into the ceiling, it starts to smash up, and more and more of the paper it at the highest level. the more paper is at the highest level, the more distortion.
ok, so that's our conceptual model of how gain stages work. you can apply that to power amps for both home and cars, and also for the pre and power amps in guitar amps (doing that in the preamp is how we get guitar distortion at lower levels these days).
the following is an example where a clipping amp will blow our speaker, but it is not an example that shows that clipping was the CAUSE of the blown speaker.
now, assume we are already at the very brink of overheating our subwoofer. we're really rocking out, and we've got the gain cranked way up. our amp is distorting like a mother, but we don't care cause we're totally ghetto. as we crank the gain up, more and more of the signal is being sent to the speakers as maximum output (not as peak, but as maximum continuous). when the gain gets really high, nearly ALL of the signal is sent as maximum output, which means the speakers are being subjected to the amp's maximum continuous capacity nearly all the time. if the speakers were overrated in capacity (i think we could assume this given the industry), then by feeding them this constant signal that exceeds their capacity, they will overheat and poop out. by having the gain really high, we ensured that the amp sent a really loud signal all the time, instead of the usual crests and falls in level. so if the speaker was overrated a little, or wasn't cooled properly, then a clipping amp could result in speaker failure, not because of the peaks (there weren't any, the caps were drained) or because or the type of signal or the harmonics, but because it was sending it too loud of a signal for too long. it is in this fashion that an amp that slightly exceeds the *real-world* capacity of a speaker would blow it. the point here is if the clipping amp killed the speaker, then the speaker wasn't rated accurately. that doesn't change the fact that Joe Blow hooked his 700 watt amp up to his 1000 watt speaker, dimed the gain, and it blew. It just means Joe Blow's "1000 watt" sub was actually only good for less than 700 watts continuously, or that the amp company was very conservative in their power ratings for the amp (less likely, why in the world would they want to do that?).
it's worth pointing out that if the amp hadn't clipped, it would have been even worse. by allowing "the top of our paper to go above the ceiling," we would have simply ensured even quicker failure of the speaker (remember that although we now would be feeding it a wider dynamic range, that range would be going to the same lows, but much higher highs, producing way more heat). clipping actually prolonged the life of our speaker. so if you hear your amp clipping, that's definetely a warning that you're feeding a crapload of power to the speaker and you should back off. the fact that your "underpowered" amp clipped is what just saved you from buying a new speaker.
now for the reason someone might think that because their amp clipped, it got louder, and thus would complain to the world that their clipping amp was the CAUSE of their blown speaker: as we noted above, distortion has the effect, in addition to adding harmonics to a signal, of compressing the signal. compression is a mental trick that makes you think things are louder when they really aren't. by putting a limit on the highest of the dynamic range, you are free to increase the level of the lows without actually making the signal any louder. people do this in the recording industry all the time. most pop records are majorly compressed (you can do this without distorting). that way the album can seem really loud while not exceeding the limits of the CD. radio stations compress their signals even further, so that you can hear the quiet parts of the music better when you're driving. the music seems louder, but the peaks stay the same. as noted above, in a subwoofer situation, as the amp begins to clip, the music will seem to get louder while the peaks stay the same. Joe Blow thinks that the clipping made his amp louder when really he was just hearing the compression effect that distortion has. if his amp hadn't clipped, it would also have seemed louder, but it would have also seemed punchier in addition to just being loud, and it would have blown his speakers before he even knew anything was wrong. but Joe's amp clipped, and eventually his speakers still blew, so without any real knowledge about the cause of this effect, he blamed it on the clipping and not on the fact that he was misinformed about the capacities of the speaker and was simply turning it up too loud.