New Speakers up to $15k - The more I read the less I know!

Disappearing speakers have minimal audible cabinet resonances. They're still there but below threshold of hearing.

Harbeth M40.2 s or TDZ 712s are well worth considering. Like any loudspeaker in existence they'll have faults for sure, but cabinet colorations won't be one of them. Both spooky good on imagery too.

Planars are also noted for obvious low cabinet resonance but I've not heard one here in the UK I could recommend. Quads just don't sound right with Pop to me. Or maybe they are just too honest for their own good! 

http://www.harbeth.co.uk/hifi-speakers/monitor-40-2-domestic.php

http://www.eclipse-td.com/uk/products/td712zmk2/index.html

@cd318 -- agree with your point about cabinet resonances, but given that I was surprised at your Harbeth recommendation.  From John Atkinson's measurements of the 40.1:

"There are no obvious discontinuities in the impedance traces that would suggest the presence of panel resonances. However, that big, minimally braced cabinet is considerably more lively than the norm; there are many closely spaced resonances detectable on all surfaces. Fig.2, for example, is a cumulative spectral-decay plot calculated from the output of an accelerometer fastened to the name plate on the rear panel. Many ridges of delayed energy can be seen between 80 and 250Hz, and while these varied in strength depending where on the panels I placed the accelerometer, there was no position at which they were absent. The point of this behavior, of course, is to make the cabinet panels floppy enough that the inevitable resonances will be low enough in frequency that they fall below the region where they will have audible consequences. This is the opposite of conventional enclosure design, in which panels are made as stiff as possible to push the frequencies of the resonances above the region where they will lead to coloration."

Apparently the speaker is designed to somewhat compensate or maybe even use cabinet resonances to its advantage, but it appears to do very little to move them to below the threshold of hearing.  Between the thinly-braced cabinet, the associated resonances from 80-250Hz (certainly not below the threshold of hearing), and the wide baffle I'm frankly baffled as to how these speakers disappear at all.  While they seem to make it work somehow, the Harbeths seemed like an odd choice given your initial statement.

@soix You are correct in that the resonances are easily measurable.

However it’s Alan Shaw’s opinion (designer/Mr Harbeth) based on historic research carried out by the BBC that the best way of dealing with these inevitable resonances is to allow them to move below the threshold of hearing. Then it won’t matter how they measure as long as you can’t ever hear them.

Harbeth argue that brute force methods of resonance control only serve to move them upwards in frequency to where our hearing is most sensitive. Any coloration in the midband, even the mildest, can quickly destroy the illusion of reality.

Since loudspeakers cabinets will always resonate due to considerable internal pressures caused by the movement of the drive units and the harm they do as they try to force their way out of the box and back through the drive unit itself, something must be done to reduce this pressure. But what?

The Harbeth approach of lossy cabinet design is one method of dealing with these resonances. Doing away with the cabinet altogether as in Open Baffle designs is another, but that has its own issues.

Still, I prefer either of those approaches to the solid 2 inch thick, glued, screwed, reinforced cabinet wall approaches where something beautiful can so easily get lost.

Can't say enough about how good I find the GoldenEar Triton Reference speakers.
Hopefully you could try them in your house if you are interested (borrow overnight from a friendly dealer).  And you'd have $7K of your budget left over...

My 2 cents on the Harbeth (and other wide baffle designs):

The Harbeth speakers - the larger ones included - do a surprising level of "disappearing" as sound sources. Not what you’d expect from looking at them. So their design philosophy does seem to be working to quite a degree.

That said, Harbeths don’t "disappear" as sound sources as well as any number of other speakers that take the "dampen resonances" approach.

Thus far I’ve auditioned 3 brands of the wide baffle, resonating cabinet approach: Audio Note, Harbeth, Devore. In each case they disappeared more than their shape lead me to expect. But also in every case, that aspect was pretty easily bettered by other resonance-controlled speakers (e.g. my Thiels and others). In the wide baffle/resonating designs, I always sense some level of "fill in" between the sonic images. This can lead to a nice cohesiveness and richness to the sound, but can also become a bit more confused with more complex material, with a lot of instruments playing in the mix.

One of my go-to tests for this are some cuts from the Lost Angeles Guitar Quartet, in which 4 similar sounding guitars are simultaneously playing complex parts. The wide-baffle speakers I auditioned could make it harder to untangle one guitar from the next, vs my Thiels and some other low-resonance speaker designs.

The trade off is that wonderful bigness and warmth of tone from the wide baffle designs. I’ve really enjoyed what I’ve heard from Devore speakers which is why I’m contemplating buying a pair from the "O" series.