"The impedance at the mid HF crossover is far more often than not a slight dip. Most people, and I usually don't, correct the impedance rise of the tweeter with frequency, so you usually see a significant rise of frequency above the high pass crossover region. There is not much power here. That is why it makes no sense to average the speaker impedance frequency curve to come up with a "nominal Impedance.""
"But, that's the way we have always done it".
This is a response I would expect from many speaker manufacturers and in fact, a lot of speaker spec sheets show "Nominal Impedance- 8 Ohms, 6 Ohms or 4 Ohms."
"In an ideal world all amplifiers should be designed to cope with this reality. However at the price point of receivers, and given the heat constraints of having everything in one box, this is not going to happen. However every speaker designer I have ever known, and I include myself, never takes amplifier current limitations into account."
Right, but you're not crimping a cap in series with the tweeter and calling it "Done". The number of speakers with this "crossover" is incredible. Are these some of the cheapest garbage I have seen? Absolutely but they sold tens of thousands of pairs and at the time, people who wanted a cheap stereo thought they were OK. Oddly enough, some thought they were great until they blew up.
I know you mentioned designing a crossover that is almost purely resistive but how many speaker manufacturers see that as economically feasible?
It's all about acceptable compromise, don't you agree?