The most salient characteristic (to me) is that the acoustic presentation of some of these speakers seemed quite forward (row D), whereas that of others was really quite laid back (row M). There was also, quite often, a second correlation between that forward presentation and a (relative) brightness in the treble.
The last sentence in this para says it all for me - what you are listening to is speaker phase distortion & lots of it. This is a classic case. It's an effect of using higher order x-overs in the various speakers you auditioned. Higher order x-overs (like 2nd, 3rd, 4th & higher) impose a large phase shift onto the incoming music signal. This ruins the phase of the original music signal & what you hear as a result is heavily distorted music where the phase relationship amongst the bass, mids, highs is no longer what was read off the program material. In all higher order x-over speakers, the speaker drivers are phase coherent at only their x-over frequency & the moment you go +/- that x-over frequency there is no phase coherency any more.
One thing to remember here - as you go thru the audio frequency range of 20Hz-20KHz there will be phase shift (due to the drivers & the x-over) - that's *not* what I am talking about. What i'm talking about is the phase *difference* between the bass & mids, mids & highs, bass & highs. The high order x-overs destroy this phase relationship & as a result destroy the music coming out of the speakers. For the music to sound correct, the speaker must maintain the phase difference between the bass & mids, mids & highs, bass & highs.
Why am i harping about the phase? Because phase is the independent variable that exists in nature. Frequency is a derivative of phase. Also, during recording music, distance of the artist/instrument from the mic is phase i.e. time of arrival of sound to the mic is phase. if you change that phase relationship, you get different frequencies. And that's what these various speakers you auditioned are doing with their use of higher order x-overs.
if you look at impedance & phase plots of speakers that use first-order x-over circuits, you will notice that the phase remains flatter over the 20Hz-20KHz region & that phase relationship/difference between the bass & mids, mids & highs, bass & highs is maintained over ALL frequencies (yes, ALL) - not just at the x-over frequencies. This is the physics of a 1st-order x-over (& not something that I conjured up). No other higher order x-over has this intrinsic/native property. For this reason, you will find that speakers made with 1st-order x-overs sound more correct.Speakers made with 1st-order x-overs are, what we call, time-coherent (which means that they are both phase coherent & time-aligned all at the same time). Speakers made with higher order x-overs are phase coherent only at their x-over frequencies & still other speakers made with higher order x-overs are time-aligned & phase-coherent at their x-over frequencies.
you wrote that the Vienna Acoustics Baby Grand sounded the best of the lot you auditioned. I looked up a Stereophile review - these speakers use a 1st order (6dB) & a 2nd order (12dB) Bessel x-over. A Bessel filter is also known as a "maximally flat phase" filter. Whoa!! Guess what? Vienna Acoustics used a x-over that tried its very best to maintain the phase of the incoming signal!!! Exactly what I wrote above.
What a coincidence then that it sounded the best!! Not a coincidence at all...... The engineers at Vienna Acoustics seem to have a better understanding that it is indeed phase of the incoming music signal that must be preserved if the music is to sound good.
IMO, speakers even better sounding that the Baby Grands would be 1st-order x-over speakers such as those from Green Mountain Audio (in Colorado Springs/Denver - they might have moved recently), Vandersteen, Sanders Sound Labs. None available in your area from the looks of your OP but you might be able to get a demo if you drive.
As far as I can tell, these features are often preferred and indeed seem to be aimed for in the voicing of many models during their development.
I don't think so but if the speaker designer cannot solve the problem one way around it is to state that this particular distortion is a wanted feature of the design. There you go - now you don't have to solve the problem anymore. Sell it to the unsuspecting public as a "revealing" speaker. total BS!!!
The only way to go is to have a time-coherent speaker. We have discussed this matter on this forum many times starting 2002 (yup, 2002!). This is an awesome thread (& many consider this the best Audiogon thread) that was started by Rbischoff in 2002 about time-coherency & it's important. It came up again very recently in the the superb thread "Sloped Baffle" - pay particular attention to the posts made by Roy Johnson (who owns Green Mountain Audio. you will probably think that Roy J has an agenda to sell his speakers but try to put that in the back of your mind & just read his posts. He does an awesome job of explaining why time-coherency is very important).
http://forum.audiogon.com/cgi-bin/fr.pl?cspkr&1403209611&read&keyw&zzsloped=baffle.
in the "sloped baffle" thread, I give the URL for that thread started by RBischoff back in 2002.
Lots of posts for you to read but very much worth it, I think.
Also, there are few members who have been converted to using a time-coherent speaker (after several years of NOT using a time-coherent speaker) by buying DEQX. And, following this wake, there are other Audiogon members who are seriously considering buying DEQX for their systems (convert their respective speakers to a time-coherent version). You can read that discussion & their experiences of being converted to time-coherent speakers in the "IS DEQX a game changer?" thread
http://forum.audiogon.com/cgi-bin/fr.pl?ddgtl&1408492712&read&keyw&zzdeqx=a=game=changer