Dani, I'm going to try to sort out some different areas of discussion, theory.
In no particular order:
-- What you're talking about with voice-leading is I think called Neo-Riemannian theory. There are whole huge dreary compendiums of the stuff. It is the going cottage industry for employing grad students and teaching assistants etc. in music theory. I just bought about five of these books, I may read them.
-- For a brief shining moment with serialism, it was as if composers could write books talking shop with other composers. Serialism made for good shop-talk. There was also a spirit of interdisciplinery research shown in the good books on music theory that were written by composer-theorists. Then that stopped, either because publishing changed, or the scene changed. I don't see books like that any more, even in serious music bookstores.
-- My approach with serialism is this: Generation of material is one thing, taste is another, hearing is another. A slightly augmented set of standard serial techniques is an excellent toolbox for generating pitch-material. It can be used to generate macro pitch sequences lasting the entire piece, or surface tunes that go by quickly. The virtue of the box of techniques is that everything is at least logically related. The composer can also insist if she wishes, that the musical results of such techniques are also always about being audibly related. I myself do insist. Others don't.
-- I've witnessed even fairly bright composers (I'm thinking of you, Daniel F.) who get bogged down in tedious and strict definitions of what serialism "must" be or ought to be "in essence". (It has to be about "interval fixing" or something.) No.
-- Instead, serialism to me as a composer is a generative technique, nothing more or less. It has to produce audibly related results. It also has to produce pleasing, musical results, which means I have to cherry-pick insanely to get the unlikely coincidences or tonal contrapuntal combinations.
-- To me, logical relations in composition are there to produce only the psychological result of hearing a resemblance. This sounds like this. It's like this, but different in this way, same in that way.
-- Therefore, mathematical invariance is less interesting than related variance.
-- Consider what happens, for example, when a melody changes from major to minor. It's an inexact transformation or resemblance when considered literally. At a higher level, I suppose it's perfectly systematic, but not from the pov of serialism.
-- Serialism was never a complete system in anyone's hands -- everyone was using some of the same techniques but making it up as they went along, too.
(Even Babbitt and Wuroninen)
-- What I'm talking about with harmony is best observed with synthesizers and computer-controlled tuning that can be set up quickly, as with Scala format.
If you (Dani or anyone else) take that path, it opens up a whole meaningful territory of harmony in 3D and living color, if you will.
-- Harmony with only 12 tone tuning is like 2D checkers. However, there are plenty of people who could and can do more with 3 notes than I can do with 60, so it's not a question of virtuosity or musicality.
-- It's only that 12-tone 2D thinking is a sort of theoretical limited Flatland. It can be a revelation to experience something more 3D.
-- Obviously for people whose musicianship exceeds their theory (most good musicians) the 2D aspect of 12 tones doesn't occur to them at all -- they're fully capable of transcending the limits of their own thinking in their music-making all the time.
-- JI theory (7-limit, 11, 13-limit) is something I sort of live with every day, but don't try to understand all that systematically. It means that intervals like 7/4, 11/4, 18/13, become real intervals, become real consonances, become meaningful.
-- I mention low interval limits because I've talked to musicians who don't seem to understand what bass is, what inversions are, what the overtone series is.
-- Heck, there was a kid at the MIT experimental studio before it became part of the media lab, this kid was from Juilliard. This kid, who had been at Juilliard for several years as a comp student, didn't know that an octave was a 2:1 ratio. This is the degree to which standard music theory as it's taught is sort of divorced from reality.
Well, top of head. See if any of this clears anything up.
. Life is like a boomerang sometimes.
