Time-of-arrival-stereophony is what naturally appears in nature. It is the most natural way of expressing the distance from the sound source to both ears. Depending on the angle and the amount of sources we will hear things in a different moment in time.
It appears when you manipulate the time of arrival using technical devices.
Level stereophony that does not appear in nature as perfect as it could be represented while panning a mix of a recording. When we have a direct sound that is mono like for example a voice, saxophone or trumpet the amplitude of panning should not be more than 25dB (between -12 and +12 dB) as it will appear the most adequate to what we would hear in real life.
Main mics cover the setting of the ensemble as it is in the recording room. A-B stereophony creates time-of-arrival stereo only.
If we record a single instrument, it is the direct signal mono miking that creates the closeness to the sound source plus the stereo that gives the space and surrounding to the sound source.
If we on the other hand record an orchestra where this technique is most common, we do it to boost the front of the orchestra which are the softest instruments and those are the strings. Decca tree in front of the orchestra is mostly just a boost on top of the all sections miking.
Coincident, directed microphones produce level differencies only.
It‘s a really cool technique to use for intruments with more witdth than a mono instrument like for example drums, vibraphone, piano etc. (but of course we can use it on mono-instruments as well as an effect) but on the other hand that are closer together than an orchestra (where a better solution is a decca tree) or to record a direct sound and the surrounding.