Back to the subject, let me offer a simple outline of what makes the Three, Four, 3C such outstanding speakers on a technical level.
Midrange - metal drivers are lower in distortion because they are more rigid and transfer more energy into the air without absorbing or shifting this energy into distortion as does paper, poly, kevlar, etc. You get lower inband distortion, but high frequency, hopefully out of bandwidth distortion goes through the roof. Most speakers use poly or paper because they are better behaved. They give you, say, 80% performance in band and 70% out of band. Metal can give you more like 95% performance inband, but maybe 40% performance out of band. The trick is to remove out of band info.
So, with the more rigid, more detailed 6.5" midrange, you need to cross it over low, below 1000Hz, not the typical 3000-4000Hz. A tweeter can't go that low, so the dome midrange takes over. This means we've isolated the best aspects of the aluminum midbass driver - almost pure clarity with vanishly low distortion compared to a typical 6" midrange in upper bass/low mids. This is enhanced by the 1.5" baffle and poured composite frame which has lower resonance than metal.
The 2" dome adds even more performance. Most every bookshelf and tower has mediocre to poor upper midrange dispersion and distortion attributes. 6.5" midranges start beaming by ~2000Hz meaning a lack of natural "presence" in sound. And since a 2" dome is FAR more rigid and light than *any* 6" driver, it is much more effective and producing resolution with virtually zero cone resonance. Cone resonance is something that plagues most speakers, but B&Ws specifically because of their high frequency, low slope crossovers that are inappropriate for Kevlar. Revel uses far more scientific crossover/driver choices, so it's not just NHT. So, the midrange of the Three/Four offers virtually unparalleled dispersion and resonance characteristics.
The 2" dome also allows for a smaller tweeter crossed higher. The .75" tweeter in use will have wider dispersion above 10khz than a 1" dome for more "air". And the smaller dome is lighter and more rigid. At the very least, this yields a higher resonance frequency. In essence, this dome will perform more like a diamond or beryllium tweeter, but without the added cost. But it has one big advantage - dispersion.
Moreover, in general, each driver is handling a smaller bandwidth than a 3-way. That means it is easier to get a blend from each driver to the next. And, since they are all of the same material (and because they are lower distortion), the blend is further enhanced. Each driver has to handle about 2.5 octaves, not 3.3. And the 10" bass driver will go much lower, not requiring a subwoofer
So, a Four tower, compared to a typical 3-way tower like the 703 will have wider dispersion and smoother upper treble, it will have far more resolving and wider dispersion sound in the upper midrange/lower treble, it will have much more midbass detail and a better blend with the bass. The bass will be deeper with higher output and a more effortless sound and "oomph". The Four goes about 1/2 octave lower than a speaker with 6"-7" woofers and sounds better doing it.
Moreover, as you can see with the 703, the center channel uses a different driver and configuration as compared to the 3-way tower. The 703 uses a fixed suspension driver (a REALLY bad design, but that's for another thread sometime) instead. The 705s match the HTM7, but not the 703. But 703 is going to vent out of phase bass in the 40-50Hz region, right in the middle of the bass range. The 705/HTM is going to vent out of phase bass in the 60-70Hz range and added distortion and noise through the low midrange. And the 703 will likely have different phase characteristics in the midrange/bass compared to the 705/HTM which will yield less accurate and natural imaging. The Three/Four/3C doesn't have this problem because the mid/treble driver complement is the same and the Fours crossover is low enough to stay out of the way of the midrange sound.
Compared to the mighty Nautilus speaker, the Four has several performance advantages. First, the .75" tweeter will outperform the 1" Nautilus tweeter. The 2" dome is a wash as they are, in theory, equal. The Nautilus midbass, however, is too small for the 12" woofer. It needs to be a 6" for the best blend. However, B&W seems to like a thin low mid/upper bass region. This is why most B&W fans LOVE female vocals for demo purposes. However, put on a male voice and it shows the problem. So, the Four has a better midbass choice for a better mid/bass transition. The Nautilus gets points for the acoustic suspension/transmission line cabinet design, but, hey, it's got to do *something* cool for the 25x price differential
In any case, the Three has idealized dispersion and distortion characteristics that no 2-way bookshelf can match and adds the improved transient response of the acoustic suspension design. The Four adds bass to this nearly ideal design and really can only be faulted for having a ported bass design over a more expensive to implement acoustic suspension one. That would have required either a large box or powered design with EQ. I am *hoping* to convince NHT to build a $2500-$3000 "Five", but we'll see. The Three does the 8 octaves a bookshelf can really handle about as well as is technically possible and far better than many $2500-$5000 bookshelf designs. The Fours do 9-9.5 octaves out of 10 about as well as is possible. I can't think of a single speaker, however, that has idealized performance in all 10 octaves. That would require a 4-way or 5-way acoustic suspension design with a very proportional and wide dispersion design. I can't think of such a critter. Not even the Nautilus really applies, though it's pretty close, losing points for the upper bass and upper treble mostly.