Nick,
On M-S Eq - Selective Front and Rear - using M-S on the whole mix tends to make the whole picture really muddy. I tend to use it for emphasis of solo passgages.
Height - this is an interesting one. About 30 years ago, around the time that early experiments with consumer Quadraphonic sound systems was underway, some interesting work was done with 3D speaker systems. Quad, once the phase coding was sorted out, went on to become Surround Sound.
However other innovations have been made/tried (most failed) - Tomita (anyone remember him?) did some interesting work - he developed a five speaker system with the 5th speaker being mounted above the listener. I attened a demo in Tokyo at the time and it was ideal for his style of mostly moog work - needless to say it did not catch on. Remember SONAB - they tried with some success to develop a true omni direction system. QUAD did wonders for the home and studio with their Electro Static Moniors. The real interesting thing about most of these systems was the attempt to optimise the phase relationship between depth and height in a 3D space. (Additionally look at all the psychoacoustic research of reccent years)
Yes any sound can be defined as an accident between the sound source and the listener and in a perfect world every instrument would have its own unique speaker and could be placed such as to mimic its position in the orchestra.
Recently attended a demo with a 22(5).1 system (17 front, 5 rear and 5 100hz ABR) - the real interesting thing about this demo was the height relationship of the front speakers, these were mounted on a arch. The matrix base processing system was capable of placing a sound more or less anywhere in the 3D space.
Most of my stuff is bland interstitial stuff for broadcasting and as such does not warrant the work required for Film or performance work - however I find the whole area of working with samples (and the problems that that presents) most interesting and the possibilities as ever are endless.
tattie