Why Aren't More Speaker Designers Building Augmented Widebanders?


Over the years I've owned a number of different speakers - KLH, Cerwin Vega, Polk, Opera Audio, Ars Aures, and Merlin VSM. One thing they all had in common was a crossover point in the 2000 hz (+ or -) range. I've read reviews of speakers where the reviewer claimed to be able to hear the crossover point, manifested as some sort of discontinuity. I've never heard that. My Merlin VSM's for example sounded completely seamless. Yet my new Bache Audio Metro 001 speakers, with a single wideband driver covering the range of 400 hz to 10,000 hz, augmented by a woofer and a super tweeter, sounds different from all of these other speakers. The midrange of the Bache 001's is cleaner, more coherent, more natural than I have heard before. Music flows from the speakers in a more relaxed manner, and subjectively dynamic range is greater, with no etch or brightness, and no loss of resolution compared with the Merlins. I have to conclude that Bache's design has an inherent advantage over more traditional designs with a crossover point or points in the midrange frequencies. I wonder why more speaker designers haven't tried this approach?
128x128cellcbern
The W3 series, both ferrite and neodymium play out to 20 kHz.  The neo has a flatter frequency response, but they’re both very good drivers.  Going back to the OP, I think one of the challenges ( at least for commercial production) with crossing so low is that the crossovers end up costing more than the drivers.  Re-capping mine with Mundorf Supremes approaches $500.00.  But, I’ve built a few speakers with the typical 2nd order L/R at 2000-3000 Hz crossover- using pretty decent tweeters- and the wide banders come out on top.  I’m not sure why—they just sound more coherent.  And despite the fact that I’m using large values for caps and coils, the parts count is actually pretty low- not a lot of parts in the signal path- which might explain why they sound good.  Some of these designs use a simple first order x-over on the woofer and second order on the FR.  With impedance equalization you’re looking at maybe 5 components in the x-over, maybe less.  Compare that to the parts count in a 3-way.  
Another good design is the Silverline Minuet Supreme. 3.25 inch woofer crossed to the tweeter @3500hz. The little woofer can play very high and the tweeter only comes on for the highest notes, delivering the airy and delicate top end that a widebander can't. Crossed over to the sub @60hz makes for a really nice small room or desktop setup. Icing on the cake is that it's easy for a tube amp to drive.  
My mistake on the Altecs! They are the 755 series, with the 755A's being the most desirable because they claim to be built by Western Electric. I had a pair of 755E's. Not as rare, but still sounding fine (and a lot cheaper!). An excellent DIY choice for a 1 driver system. They particularly excell in the midrange!
And in a two- cubic foot box they have quite satisfying bass - and high sensitivity, too!
I have a Feastrex Full range driver, 9inch field coil. It practically covers from 60 Hz to 18khz. I augment with a pair of subwoofers (REL) and Supertweeters (Townshend Ribbons). I agree with cellcbern that the midrange of widebanders offers something special compared to conventional multiway speakers and that's why I've stuck to fullrangers/widebanders for many years.
I think the issue is not just the crossover point but also the phase. Additionally you can't ignore that below/above a crossover point, the music is handled by different drivers with different impedances, sensitivities etc. Hence, its hard to get music to sound coherent as it crosses the crossover point frequency. 
What you gain in coherency though you give up in other areas e.g. frequency range, dispersion, power handling and hence dynamics etc.
But for me, I've found the strengths of fullrangers/widebanders to outweigh the negatives. YMMV though.