Magfan - It cannot be done since woofer midrange and tweeter have completely different power ratings. Applying full power at midrange frequency will most likely damage midrange speaker and doing the same at high frequency will definitely take out the tweeter. It is difficult to estimate proportions between lows, mids and highs since it is recording dependent. Music power delivered to speakers is only a few percent of peak power so measuring speakers' max RMS power is not the answer.
Russ - You have to remember that listening is logarithmic so power has to increase/decrease ex-potentially. Half as loud means 1/10 of the output power. If you add to it silence in the music it will become obvious why average music power delivered is only few percent of peaks. Measuring amp with sine waves (test tones) on a dummy load won't tell the story either since speaker impedance is complex. In addition 100W rated amp might put much higher momentary peak power - design dependant. IMHO measuring doesn't make much sense. I would be the most afraid of under-sizing amp since it will lead to clipping = high frequency energy delivered to tweeter = overheating. If your amp is strong you should be able to hear when your speakers distort.
Russ - You have to remember that listening is logarithmic so power has to increase/decrease ex-potentially. Half as loud means 1/10 of the output power. If you add to it silence in the music it will become obvious why average music power delivered is only few percent of peaks. Measuring amp with sine waves (test tones) on a dummy load won't tell the story either since speaker impedance is complex. In addition 100W rated amp might put much higher momentary peak power - design dependant. IMHO measuring doesn't make much sense. I would be the most afraid of under-sizing amp since it will lead to clipping = high frequency energy delivered to tweeter = overheating. If your amp is strong you should be able to hear when your speakers distort.