Howdy, why would this eliminate SBIR?
They don't eliminate SBIR.
SBIR improvment is one half of the advantage however. You effectively halve the mess since you are only radiating into a half space. Since the worst of the mess comes from the nearby surfaces removing the rear wall is a big win. Once you get to second reflections you are already well ahead of the game, and the mess remains devoid of the strong effects due to the first reflections.
The second win is removing baffle diffraction effects. No more baffle step compensation, and the mid range is no longer controlled by proximity to the edge of the cabinet. We tend to be oblivious to this part of the equation. The reason the speaker measured here has such a wide directivity well into the upper mids is simply that there isn't a baffle edge reflecting the energy back. What you see is the natural directivity and response of the driver. Choice of crossover points remains important but is directed more by the driver diameter and capability than the baffle design.
In ordinary cabinet speakers the designer is always compromising things in order to get a useful response in the face of the cabinet effects. You don't see this in any obvious place, as it gets folded into the overall driver response, but it is sitting there under the surface and is responsible for a whole host of second order ripples and artefacts.
Clearly the design response needs to take account of the different mounting, as the final at-listener response still needs to be level and smooth. But this becomes easier.
Can someone explain why resonances are not more of an issue? I would assume most are installing into drywall and resonances would be a huge deal. Also the cavity is often shallow (back reflections) and it's volume could be limited by fire breaks in the walls (horizontal 2x4 between studs).
Drywall and other walling materials are usually fairly well damped. You don't notice them resonating in a room with box speakers driving the room, yet the energy hitting the walls is similar, just more spread out. Indeed drywall is a good starting point for a dead walling material.
Box resonances don't need to be any bigger problem than any other sort of box. The dimensions are a bit odd, and so the frequencies that resonances might appear at are different, but other than that, nothing special. The very shallow depth tunes the F-R resonance out of the passband of a woofer. If you have a 2-way you might have a bit more to contend with, but some appropriate absorption will help. IMHO such resonances are often unappreciated in importance, but they are not insurmountable. If you see peaks at about 800Hz you might guess where they came from.