I'll be putting the surround files of this concert somewhere in due course, but not straight away.
I get the operation of a binaural mic, I suppose, with the mics aimed away from a center. But a surround sound mic? Is that set up for right and left of the choir and right and left of the reverb?
The usual way of recording surround is somewhat as you outlined - and is no more defined than what you wrote.
What I do (ambisonics) is different. An idealised description (the theory of why this is insufficient as it stands is beyond the scope of this margin) is that the microphone is measuring the
pressure changes at at point (that's the same as an omnidirectional mic recording a mono signal) and also the
movement of the air as three orthogonal velocity vectors (equivalent to the output of three figure-of-eight microphones). The decoding into loudspeaker signals attempts to drive them so that the original pressure changes and movement are recreated at the centre of the loudspeaker array. You can easily object that (1) the head is larger than a point, and (2) a point is in any case an impractical area of reproduction for multiple listeners - however, in practice, the simple theory outlined above is not all there is, and it actually works out pretty well.
Some of my older recordings were made with a microphone assembled from an omni and figure-of-eight capsules; but they can't be made coincident, which the theory requires, and so better results are got by using four cardioid or sub-cardioid capsules placed close together on the faces of a tetrahedron and matrixing them to generate the required signals. In fact, the patterns of the derived mics are typically better than can be obtained from physical capsules which claim to have those patterns!
Since many people (including recording engineers) don't get the idea of coincident mics for stereo, they are even less likely to get it for surround! However, there is solid theory to justify the methods I use in preference to any alternatives; and one of my aims is to generate plenty of examples of practice to go along with it - which is why some of my recordings have been used in talks and demos at AES conferences and the like.
Note that what I have described is truly three-dimensional - the practical case of surround in a plane is already a compromise compared with the whole deal. Oh, and this theoretical framework, and the microphone with which to record it has been available since a couple of years before 1980.
