After hearing a college lesson on Pierre Schaeffer and his potentiometer d’espace, Raven Sanders, a student of the Lyle School of Engineering at SMU, decided that Schaeffer had the right idea when it came to audio mixing – a 3-D approach.
Having had much experience with multi-channel sound diffusion, I believe that some control is lost with your choice of control mapping. Ideally, one should be able to place sounds in the 2D space at will without having it tied to volume. Your control scheme does not allow one to put a very soft sound in the center of the field of perception. Having the loudest sounds at the center where one has the least control over fine angular position is also not optimal.
I would suggest that volume be controlled independently by a slider to one side of the bullseye for the spatial location. That way one could fade in and out a sound anywhere in the sound field. For multiple sounds, use a different shape for each sound, then use color coding to indicate its intensity. Multiple sliders could be used for each sound, but it may be possible to map finger pressure as a 'blob change' to control the volume of the sound directly on the sound field map. There are commercial adaptations of surround panning that may be of help, for example:
Also, note that there is much DSP work that needs to be done to give the audience a convincing rendering of a particular sound in an exact location. You may need more screen widgets to control those features.
Schaeffer originally used 4 stationary induction coils, arranged just as his speaker positions, that picked up his source sound from a fifth coil in held his hand. True magnetic induction allowed him to place the sound anywhere in the horizontal plane using the front left, right speakers and the rear speaker. The 4th speaker was mounted at the ceiling so that 'up' could be encoded, thus giving the audience a 3D sound experience.