Loudspeaker dispersion-

Just looked at speaker details showing 90 degree horizontal; but only 5 degrees vertical. The dome and cones are all round. Why isn't the dispersion the same in both planes? Forgive my ignorance. I expect faster answers on this forum than trying to research this.
Both speakers work together to place images in space, between the speakers. Vertically, each speaker works by itself.
What are the speakers? Most panel speakers (e.g., electrostatics, planar magnetics) have very limited vertical dispersion, which is why they make them tall. Some speakers have wave guides that determine the dispersion pattern. A tightly spaced vertical array of speaker drivers (such as in a dynamic line source speaker) has very limited vertical dispersion.

D'Appolito (and most other) MTM arrays have a limited vertical dispersion as well. There's something about the tweeter with midranges both above and below that limits the vertical dispersion. In many cases limited vertical dispersion is intentional to minimize the sound bouncing from the floor and ceiling and smearing the image.
That's a good question!

My guess would be it has something to do with tweeter location on the front baffle and how sound is reflected off that accordingly.

Also possible that different soft dome designs have different internal dispersion patterns by design despite the outward symmetrical appearance of the dome.

Or both.
Post removed 
04-04-14: Bob_reynolds
Here's my guess...

Put one driver on a baffle and you should have the same dispersion horizontally and vertically. Place a second driver vertically adjacent to the first driver and vertical dispersion will suffer.

Yes, and doubly so when the tweeter has mid/woofers above and below the tweeter as in a D'Appolito or MTM array. tHe larger waves of the midrange drivers keep the shorter tweeter waves from having much of a vertical dispersion at all. MTMs have a famously narrow vertical dispersion. An example is the Atlantic Technology AT-1, near fullrange MTM tower. Its measurements showed a vertical dispersion of 5 deg. above the tweeter axis and 10 deg. below. The commentary also mentions that crossover design can influence dispersion, especially suckouts.

A pistonic driver's dispersion depends on its employed frequency range relative to the piston's diameter. For example, suppose we have a 2-way monitor with a 6-1/2" woofer and 1" dome tweeter, a pretty common configuration. A 6.5" dia. diaphragm starts beaming at about 2100 Hz, whose wavelength is 6.457. Let's suppose the speaker designer chooses a crossover point of 2.5Khz to improve power handling. That pretty much guarantees that the dispersion will narrow significantly at and near the crossover point.
Post removed 
That's not the wide band dispersion, that would be above a certain frequency.
As an aside this talk confirms my belief in the good cabinet design (snail) of the B&W flagship Nautilus Loudspeaker. It certainly minimizes speaker cabinet interaction, doesn't it?
I'm not sure that the baffle acting as an extension of the driver diameter could explain this. If the cabinet is one foot wide (like many are) then horizontal dispersion would be impacted at app 1khz, but the 90 degree horizontal spec cited by the OP would seem to indicate that this is not the case. Even if this speaker is really narrow, I doubt that many foot wide columns see the horizontal dispersion of (effectively) 12" tweeter.

Although the idea makes sense intuitively, I doubt that the baffle width explains the spec. My guess (just a guess): It's probably destructive interference as cited by Bob in his first post and expanded upon by Johnny.
Johnnyb53, I've read elsewhere the same comment about the relationship between the diameter of a driver's diaphram and the wavelength at which the driver will start to beam.

Just thinking out loud here .... Most tweeters have 1 inch domes or diaphram. At what frequency would that type of tweeter start to beam?
Just thinking out loud here .... Most tweeters have 1 inch domes or diaphram. At what frequency would that type of tweeter start to beam?
As the frequency approaches the diaphram diameter, the dispersion narrows. In the case of the 1" tweeter--barring some kind of waveguide--it's in full beam mode at about 13,500 Hz. See the wavelength caluculator to find the wavelength of any frequency.
Johnnyb53...thanks for that link -- most instructive website!
In most cases, speakers use drivers lined up vertically, the dispersion of the individual drivers doesn't vary, but the summed output will vary because the drivers are not concentric but positioned in a line and putting out different frequencies.
I appreciate the summed output comments. However a 90/10 ratio is extreme. I hope some designer might "chime" in about this. (Are chimes directional?)