@farmerman,
MY humble opinion, you should restrt nd try a more simple approach that would actually be more genesis oriented. Like , across the x axis have the actul amount of free silica, y xis from the top, hve the feldspar/feldspathoids (melanocrats would "fall out". THEN, for each x/y group, split that into plutonic v effusives (Stuff like alaskites nd porphyrys would plop out)
THIW WOULD, of course, only be for igneous rocks.
The same can be done for metamorphics except they would be discussed bse upon (Chlorite to kyqnite/ sillimnite gradation) as well as presence of specific garnets.
BTW, such charts have been available for many years. I think you are trying too hard to make a "Mineralogical theory of everything", no?