At 82 to 86 dB...loud, but not ridiculous.
SW X/O at 40 Hz...14 vrms and 21 volts peak.
which is 49 watts rms and 110 watts peak.
SW X/O at 100 Hz...12 vrms and 18 volts peak.
which is 36 watts rms and 81 watts peak.
Note the following facts.
1...SPL measured at middle of room about 12 feet from speakers. But, with Maggies, the SPL changes very little as you move away from the speakers.
2...SPL measured with SW operating. I don't have a convenient way to drive the Maggies full range (balanced/unbalanced interconnect issue). However, since Maggies roll off quickly (MG1.6 at 40 Hz) you might as well roll them off electronicly even if you dont use a SW. My findings apply to setups that do use a SW.
3...When the X/O was increased to 100 Hz, the highest rms voltage readings did not come down much, but most of the time the rms voltage was significantly decreased, to 3 or 4 vrms. This suggests that the peak power requirements were related to bursts of sound higher than 100 Hz, and so unaffected by the higher X/O frequency.
4...Music was a DVDA played in stereo, Emmylou Harris, Producer's cut, track 1, "If I could only win your love". Selected because there is limited dynamic range. (SPL doen't change much).
5...Radio Shack SPL meter. Digital readout. No guesswork.
6...All this is for MG1.6. Maybe someone having MG3.6 can contribute another test.
The following are opinions...
1...I suspect that the maximum short term (1 second) rms voltages were, or with different music could be, higher than the values that I caught with my voltmeter.
2...I suspect that the short term peak voltages could be higher than 1.5 times the rms.
Conclusions....
1...If you use a subwoofer, and dont play Sousa marches or Wurlitzer theatre organ performances, a good 150 watt tube amp with good clipping recovery characteristics will work.
2...250 watts would be better.
3...My 600 watt amps are not overkill.
4...Sadly, the little inexpensive "Six Pac" triode amps, 55 watts, are out of the question for me.