OK, now...a simple question: If I were using this system and I wanted to invert, say, an envelope. Not an uncommon thing; you see this quite a bit with EGs for VCFs, like on my Jupe-6. Soooooooo...where are the attenuverters? Yeah, I know...attenuators and the like are dull and boring and they've not got all the blinky lights and crap, but try using the "big stuff" without them and you'll discover the exercise in frustration that's waiting for you here.
Also, the idea of having SIX voices AND percussion in a system this small is...uhm...ambitious, but in a build this size, that's a REALLY BAD idea, since you're either going to have to cut corners (and this is where a lot of new users majorly screw up) and remove important things, or do the sensible thing and realize that you simply shouldn't try to put that much in just 352 hp.
I would suggest losing ALL percussion in here, because it's far cheaper and, in truth, less frustrating to simply use a standalone drum machine that's locked to a central clock. That'll open some space. Then I think you might want to swear off that six-voice architecture; sure, you COULD do that, but your likely results will either result in six badly-implemented voices, or...well, something like this, where there's very few "helper" modules even though the voices are well done. You're simply going to have to scale this back. Also...
"I have purposely built separate voices to keep me disciplined somewhat, so that I stick to certain voice structures and combinations.. at least until I get familiar with everything.."
A better method to learn your system would be to put it together gradually...after, of course, correcting what's here already. By slowly building up from a very basic module set to the intended goal, you'll gain a much better appreciation for the instrument and also, you'll realize that there's not exactly such a thing as a "voice" in modular in the first place. It's not fixed in that way...but it's very open-ended, so if you wanted it to all be ONE "voice", you could do that. Moreso, you can have different "voices" interacting and affecting each other, as you'd find in generative works.
One last thing about VCAs, also...there's TWO types. The ones you usually see as "add-ons" in modules tend to be basic linear VCAs...but we don't perceive apparent loudness as a linear function. The decibel scale is actually logarithmic...works in powers of 10 so that +10dB is a doubling of the apparent loudness, which is how you get to the Threshold of Pain so quickly. Anyway, those linear VCAs don't work like our hearing apparatus, but EXPONENTIAL ones actually do behave in ways that our hearing recognizes as changes in amplitude that one would find with acoustic instruments. Hence the little per-channel "shape" knobs on Intellijel's and Mutable's quad VCA modules, as well as any other derivation of Mutable's Veils. If you're using the VCA for modulation or early in the signal path, you'd turn that to linear. But for audio, you've got the exponential VCAs, so you turn that "shape" knob to the other setting. But the real fun is in the fact that you can change their behavior to cause the VCA to do something that shouldn't happen in acoustic instruments, or you can have one of those VCAs handling modulation (linear) and right next to it, two handling a stereo audio feed (exponential). And sure, you could get some module that has all sorts of extras in them like that...but can you REALLY control all parts of it, or has something been "kludged" so that you can't...but it makes things "simpler" (and ultimately, it doesn't). Keep this in mind when reworking this.