Just a few observations I made considering the tone matrix. Evidently, I didn't really do any in-depth looks into the causes behind the consonance/dissonance patterns or other probability appearances from likely user-choices. I didn't look into rhythm considerations that are necessary (looking at the 'j' columns).
Evidently the octave skips also poses a consideration, as M9 has a different effect than M2 when overlayed with 4ths and 5ths, etc. Also, I only dealt with combinations of notes X and Y in pairs. Obviously, further chordal analysis of the pentatonic scale with different chord voicings might be something to look at.
Also, feel free to fix my math. I could easily be wrong.
Or the original here: . I can take a better picture/scan if needed or impossible to read my rushed handwriting.