Is the most efficient speaker the best speaker?

Unsound, you've asked a lot of questions and it may take a bit to answer them all.

I think first it needs to be understood that human hearing/perceptual rules are understood much better today than they were 20, 30 and 40 years ago.

I may have suggested this before, but if not, the first thing to understand is how we perceive volume. One of the most important indicators to the human ear are the odd ordered harmonics, the 5th, 7th and 9th. Now that is easy to prove, all you need is a sine/square wave generator, an amplifier, a speaker, and a VU meter. You listen to the sine wave and set the VU to 0 VU. Then you switch to square and as best you can, set the speaker to what seems to be the same volume. You will find that to do so, the reading on the VU meter will be a good 20-25 db down. Other members of the 'gon have done this in the past, FWIW. Square waves are composed entirely of odd ordered harmonics. This simple test shows that we are more sensitive to them than we are the fundamental sine wave at over 100 times more energy.

Now we have known since the 1950s that the addition of loop negative feedback enhances odd ordered distortions. I think I have pointed you to Norman Crowhurst's work on the subject in the past, if you have not read it, it would be a good idea to do so now because otherwise its like you missed a lesson in school.

http://www.tubebooks.org/technical_books_online.htm

scroll down a ways, there are 3 volumes available as a free download. Pay particular attention to the chapters on negative feedback and the methods he uses to chart the Nyquist diagrams!

People have ascribed a lot to the sound of tubes and transistors; but I am here to say that the the very audible effects of 'bloom' that is a common audiophile term is really the effect of distortion. The 'brightness' of transistors is also an effect of distortion- in both cases this is easy enough to prove as you can put the amps on the bench and not measure any differences in frequency response, yet the effects can be heard on many speakers.

The ear treats distortion as frequency response variation. Its that simple. So if you want the presentation to be neutral, the amp can't make distortion. Now if that were the case, then I would have no issue with the voltage paradigm as it would then insure flat frequency response. But the reality is that amps *do* make distortion, so the Voltage Paradigm will fail at that goal.

Its my position that the effects of distortion are often more audible than frequency response variation. Anyone familiar with speaker design already knows how important the room is in any system, and how the speakers are often designed to work in a room. It is the room itself that guarantees that flat frequency response will not be realized, but our ear/brain system has a means of adjusting to the acoustics of the room. You might want to read some of the works of Dr. Earl Geddes on this one.

So- I have given you some homework. BTW, don't discount Crowhurst simply because he was writing in the 1950s. A good deal of the technology you hear today is based on that foundation.

The Thiel's often times do have elaborate 1'st order cross-overs, but not because 1'st order cross-overs are inherently elaborate unto themsleves, just the opposite is true. Thiels cross-overs are more elaborate because they take into account things like box and driver concerns, as well as to minimize impedance swings. Combined with suitable multi drives it allows the speakers to provide more of the available music in a more accurate, smoother fashion. With high powered amps readily available, I think the efforts are worth it.

JohnK, as my post was clearly addressed, I was responding to Atmasphere's post. Once again, you drag your professional agenda into the mix, ignoring the title and content of the thread in an effort to discount my contributions to this forum.

I've read that the human ear does not have flat response but rather is most sensitive in certain range of the audio spectrum associated with brightness.

Atmasphere, is this what you are referring to when you talk about hearing rules?

If so, seems to me that this is then a natural artifact of human hearing. It makes sense to take that into account when designing audio gear it would seem to me, but I would think there are many ways to adjust to that. It either sounds good or not. There is not a single technical approach that can accomplish this it seems to me. A flat frequency response would still seem to be desirable in order to allow our ears to hear the way they are designed without a bias introduced in the signal. Chose your distortion. It is always there to some extent.

Atmasphere, thank you again for your considered response. I like most others here, don't have access to such equipment. If this data is so readily available, why isn't it routinely being published?
I might not completely understand what your suggesting, with re: to speakers and square waves, but to my knowledge the only readily available commercial speakers that can accurately reproduce square waves would not be considered "high efficiency". You still have not established at what exact amounts or degrees that we find these distortions audibly objectionable. If they are measurable, regardless of how tiny they may be, we can blow the graph up to demonstrate comparisons on just how relative it is, and then verify it.
I did read the link you previously provided re: Norman Cromwell and within it Mr. Cromwell suggests that some feedback can be beneficial, and goes on to caution not to overuse it. It would appear to me that vast majority of amplifiers with any intentions of appealing to the "high end, high fidelity" marketplace follows this advice.
One can not automatically extrapolate that the differences one hears in speakers with amplifiers that measure the same is due to amplifier feedback. There are too many other possible causes.
Let's see the graphs demonstrating these distortions, and proof dismissing other possible causes, rather than assuming they are the root cause.
It is my position that, depending of course on the degree of distortion in frequency response, and we are talking about loudspeakers here, where there are gross differences in frequency response, frequency response distortions are more audible.
No argument with re: to room response, but that is something all together different than efficiency or inefficiency unto itself. I have read some Earl Geddes ideas, but there are too many examples that don't seem to fit into his theories.
Though the burden of proof wasn't mine, I already completed your homework assignment, and it didn't validate your claims.
It would appear to me that your asserting that the differences that you described as being measurably tiny, though in your opinion ultimately important, in amplifier distortions, and the ease in which a loudspeaker can accomodate such as amplifier, swamp what ever differences in frequency response linearity and range, dispersion charateristics, waveform fidelity, and self made noise other wise inherent in loudspeakers. I just don't hear it that way. I hear it just the opposite way. IME, less efficient speakers tend to conform better to the way I hear things, even if those speakers are mated to an amplifier that might or might not have small amounts of feedback. Interestingly enough, many who have efficient speakers that might more easily mate with an amplifier without feedback choose amplifiers with feedback. That is not to say that I am endorsing feedback in amplifiers, just that compatibility with an amplifier lacking feedback is not an important enough criterion alone to choose loudspeakers by. Furthermore, even if it was, one could conceivably still use such an amplifier with inefficient loudspeakers.