A very wide omni spacing can achieve that aspect of "stereoness" and envelopment down low. A wide spacing also decorrelates the hall ambience and reverberance down to a lower frequency, which makes things sound more "open and airy" and reproduces those sounds diffusely, throughout and even outside of the playback imaging space. That's a different kind of stereo thing than left/right imaging and doesn't relate to Williams SRA curves. It's not about accurate left/right imaging directionality or orchestra angle.
But more narrow omni spacings are usually going to have a more solid center, and often may be more accurate in terms in left/right direction imaging of the sound source locations as laid out on the stage, so the SRA curves then become applicable, at least for those imaging aspects.
(Always exceptions for TS taping, and here's another one- "section recording" by the soundboard at an outside event, often places band/PA at a distance where the "orchestra angle' and appropriate SRA is only 40 or 45 degrees wide total. So even in terms of following the William's curves for good L/R imaging using a single pair of omnis -and we need to extrapolate those curves, because as mentioned this scenario is a total oddball in the world of recording scenario which was never considered by the Stereo Zoom for typical studio or classical recording- we end up with Stereo Zoom suggested spacings of more than a meter. That's a spacing of more than a meter not with the intention of optimizing those "wide spaced omni" stereo aspects, but rather optimizing the "near-spaced omni" stereo sense of accurate L/R imaging in terms of the a Stereo Zoom)
So if only using one pair of spaced omnis, determining the optimal spacing between them partly becomes a question of balancing the left/right directional stereo imaging aspects against other stereo aspects.
..Or another approach is to introduce more microphones, as a way of working around the compromise imposed by those contradictory things which are pulling the microphone spacing in opposite directions.
I can setup very wide spaced omnis which providing an open, lush ambience, envelopment and "stereoness" down low, and get good center balance and directional imaging by introducing other mics. The "other mic's" could be a third omni in the center. That's the simplest arrangement - and it fixes the problem in two ways: First, the over-wide center 'hole' is plugged by introducing a center channel, which in 2ch stereo playback gets panned center, feeding both Left and Right channels and thus producing a strong mono component in the stereo signal, filling the hole; 2) There are now two stereo pairs instead of just one (actually 3 interacting pairs instead of just 1) and the spacing between the these pairs of omnis (L/C, and C/R) is now half of what it was for the former single L/R pair. In that sense, adding a third omni to the center is equivalent to halving the previous L/R omni spacing, and by doing that it's not surprising that the "hole" in the middle problem is fixed simply due to that radically different spacing alone. That in turn allows me to space the original L/R pair considerably wider without a hole-in-the-middle problem, and that gets me more of the other kinds of "stereoness" I want, which comes in addition to the improved left/right directional accuracy of three closer-spaced omnis.
A SBD feed counts as additional mics, and works similarly to a third omni in the center in the sense of "filling the hole", allowing a wider spacing without introducing problems. Using a directional mic in the center instead of an omni is a bit more like a SBD feed because there will be more direct sound and less room reverberance in that center channel. Using a coincident stereo pair in the center provides left/right directional information without phase-difference info that could complicate things when combining feeds (a coincident center pair is similar to the SBD feed in that it's L/R imaging is level-difference based without phase-differences, like much of the stereo informationpresent in a SBD feed- all pan-potted mono stuff, except maybe stereo verb, synth pads, stereo mic'd leslie cabs, etc). Using a highly directional center mic as the mid in a M/S pair provides better isolated up-front center, plus control over L/R directional imaging in the center by bringing up in some Side, and that sharp imaging up-front stereo center can rests comfortably in the lush, wide-omni stereo bed. Tight sharp accurate imaging "stereoness" from the M/S coincident center mic, plus lush open enveloping big deep and ambient "stereoness" from the wide omnis. To my way of thinking, those are logical progressions in "adding a center mic" to a pair of wide omnis.
Deca tree is just a third center omni moved forward a bit. Outriggers are just an additional pair of very wide omnis, added for the same "other-stereoness" reasons discussed above.
Every form of engineering is about juggling the imposed constraints to achieve a desired result. Audio recording is no different. Getting all the desirable sonic aspects well balanced is a big challenge using a single pair of mics. That's no different if using an omni pair or a cardioid pair. The addition of other mics in the main array, out-riggers, spot mics, section mics, SBD matrix or whatever, are all ways of working around the constraints imposed by the limitation to two microphones.
And as soon as more than one stereo pair of microphones is introduced, the situation immediately turns into that "system-working-together" thing. It's no longer an ORTF pair. It's no longer a pair of spaced omnis and doesn't play by the same rules imposed upon recording using a pair of spaced omnis alone. And we shouldn't unnecessarily constrain our thinking about using microphones in arrays by trying to conform to the compromises that work well when using a single pair of microphones.
[Editing done after posting a mess. Refresh, refresh view]