bdp24 said, "[Duke’s] approach of making speakers that are built to handle the demands of a full-range signal ignores one of the major benefits of high-pass filtering---removing the low frequencies from not just the speaker, but also the amp driving them."
Such was not the case in my experience with the Quads and Gradient subs. The net benefits of not using a highpass filter were greater.
And please note that my approach of making speakers that are built to handle the demands of a fullrange signal gives users the OPTION of using them that way. It is not mandatory. In a situation where the amplifier is the limiting factor, the greatest net benefit may well come from reducing the load on the amp. I’m not unaware of this issue, which is why I specified that I try to build "enough efficiency" into the speakers. An extra 6 dB or so of efficiency arguably makes a greater contribution towards easing the load on the amp.
"Regarding Duke’s finding of reduced transparency as result of high-pass filtering of main speakers, if a 1st-order filter is sufficient, a simple passive one can be installed on the input jacks of the amp powering the speakers, just a single cap usually."
In some situations that can work very well. I used that approach in a system I built back in the mid-80’s. The effects of a first-order highpass filter extend well into the next octave up, which may or may not be a good thing. Again what I do adds an option; it does not remove other options.