I didn't even know this thread was still going.
Obviously no one bothered looking up any past threads on this topic.
When an amplifier begins to clip, it's output power raises rapidly. A fully clipped amplifier will typically double it's rated output power at .1% thd.
That being said, a 100 watt speaker can burn up off of an 80 watt amplifier as the amplifier can deliver in upwards of 160watts when fully clipped.
If a speaker is rated for 500 watts rms power handling, an 80 watt amplifier (80 watts rms from 20hz-20khz @ .1% thd at 8 ohm), even when fully clipped should not cause any damage to the speaker, unless of course the power ratings were not correct to begin with

Sure, it will sound horrible, but their should be no ill affect on the speaker. Distortion in and of itself does not fry speakers. It is the added power the amplifier produces while clipping that fry's speakers.
The reason people fry their speakers by "underpowering" is because the amplifier of say 80 watts rms per channel is halfway clipped most of the time (don't laugh many people actually prefer this sound in their music) resulting in somewhere near 120 watts rms or more being sent to their 100 watt speaker. This amount of power, over time, overheats the coil. It could also be attributed to the fact that many loudspeakers simply inflate their rms ratings. Combine the two and you have disaster on your hands.
A true rms rating should be x amount of input power sustained for y amount of hours. I believe JL Audio's rms ratings for their speakers are typically for an 8 hour period. This means they will handle the amount of power claimed for at least 8 hours. Some manufacturers use a 1 hour sustained method, others 1 minute. The shorter the time, the higher the number, the greater the chance for failure as it is reached or exceeded. The actual individual driver design weighs in pretty heavy here too but I am not going into it that far.