Yes I agree with much of this.
A speaker manufacturer friend of mine who designs active speakers has the ability to monitor and report out when the amps clip in his speakers. As part of some tests he did to figure out how much amplification was needed, during prototyping, he tried different amp modules (same manufacturer but different output levels). The tweeter is waveguide loaded on a large 8" or so waveguide, but uses a conventional dome tweeter. At fairly normal listening levels, the amplifier clipped at least 1 time 100% of the time for each track he played. It clipped on average 5 times per track, with some more dynamic tracks clipping in excess of a dozen times. This particular speaker was not unusually inefficient, if it were a passive speaker, its reference sensitivity would likely be around 88-90dB or so. He ended up needing to go with quite a bit of power, I forget how much, but something like 1250 watts per channel total, I believe 250 on the tweeter, 500 for each additional drive. It was quite a lot of power, most people don't have that, and yet this was what was needed to avoid clipping. This speaker is still VERY limited in maximum output. About the same as what Dennis' speakers could do. So I think it is highly likely amps clip a lot more often than people realize (my own experience has been a few real world examples of receivers audibly clipping with fairly efficient speakers). It doesn't matter if, on average, you only listen to 1-2 watts, what matters is that .1% of the time when you need all 3,200 and don't have it.
