i have used the acousmodule plugi-in before with audiomulch and the works well.
The ambipanner I have referenced to works on a slightly different principle and coded to be used with a 4 speakers rig (4 speakers placed in the 4 corners of a square.
It creates a more organic representation of a soundfield.
With some changes in teh code it can be then used for different type of speakers configuration including typical dolby surround speakers arrangements...but this need additional work.
There are ambisonic decoders coded in Max, one example: http://www.grahamwakefield.net/soft/ambi~/index.htm
but I am not sure if they will work for live...
Probably this is something can be discussed on the M4L section?
Just to bring to the attention of the more experienced people on M4L programming.
Let's keep the discussion open!!