These documents below show what kind of presentation various setups gives. The recording angle and the angle between the microphones are two different things. Each document have the recording angle on top, "60/90/120° Klangkörper". That is the width of the orchestra/stage/area of the musicians etc. The diagram below shows the presentation of the recorded sound between the speakers set at +-30° (44-46° is better in most cases).
http://www.sengpielaudio.com/Stereo-LautlokEines60.pdf
http://www.sengpielaudio.com/Stereo-LautlokEines90.pdf
http://www.sengpielaudio.com/Stereo-LautlokEines120.pdf
In the diagrams you can compare a pair of cardioids (Niere/Niere) with different angles and spacings and the ORTF (halvway down the pages). You can see that the some setups results in "a hole in the middle". Others gives a very narrow presentation.
ORTF is suited to a maximum recording angle of 90°. A pair of cardioid 20-25 cm apart and with 90° between them or a pair of figure-of-eight with 80°, gives a similar presentation.
I had downloaded these years ago but had forgotten about these useful diagrams. Keep in mind that they say nothing about other important aspects of sound capture such as the nature of the reverberant pickup of the room outside the 'Klangkörper' angle (or in Stereo Zoom terms, the Stereo Recording Angle or SRA).
If very close to the stage/musicians, look at the recording angle of 120°. It seems that XY at 90-135° or AB about 20-30cm apart would be the better choice. Avoid figure-of-eight, DIN, NOS, ORTF or widely spaced AB (mostly omnis).
Far from the stage/musicians, look at the 60° diagram. Figure-of-eight, DIN, NOS or ORTF seems to give similar results. AB spaced 40-50cm can also be used. XY or widely spaced AB (mostly omnis) should be avoided.
Very close and very far is where the applicability of this data to spaced omnis falls apart in my experience. I've found that spaced omnis on stage can work very well with significant distances between mics of around 4' or ~1.2m, probably because there are intensity differences introduced due to the close proximity to the sound sources. I've run wider spacing on stage with additional mics to fill the center, but 3'-4' seems pretty safe with no hole. Spaced omnis on stage or stagelip can also work well to balance the aparent closeness and timbre of each source, since each of the players are more likely to be a similar distance from the mics (another aspect not covered in the PDFs)
Likewise for distant mic'ing, I commonly use a 3' (1m) A-B spacing at outdoor amphitheater type concerts from typical taper's section FOB distances. By my calculations that spacing gives an acurate SRA of around 45 degrees, which is often close to the angle between stacks at that distance. Additionaly I suspect that the strange situation of recording two widely spaced speaker sources that are reproducing primarily mono information (the typical concert situation) is somewhat of a special case, different from individual sources spread across a stage and so less suseptible to wide A-B spacing problems.
In my opinion, these diagrams are best applied to cardioid-type pattern configurations. Thanks for posting the links, Roger.