Just passing through, and I've never attempted polyphony in a modular, so take this with a grain of salt.
Basically, I'm getting kinda mixed vibes on this setup. The only reason I can think to have four identical oscillators is that you want to have single-input instrument / multiple voice polyphony, e.g. an analog polysynth. From that perspective, it seems like this has a ton of redundancy. My understanding of the block diagrams of most analog polysynts is that you'll have your multiple vcos fed into a mixer, then have that fed into your signal chain of vcf/eg/vca, etc. from that view of polyphony, it seems not only kinda weird that you'd need a vca (and a summing vca, at that) + eg + vcf + toolbox per channel. if that's where you want to start (keyboard poliphony), you can probably get by with 4 vco, some subset of mixers, 1-2 of the dual vcas, vcfs, and egs. basically, this could probably be chopped into a 6u. it also seems to me--again, having zero true modular polyphony experience--that it'd be the hugest pain in the ass to get four full, independent voices to sound anything similar to coming from the same synth, but that just may be me being a crybaby. for example, i could be wrong, but it looks like the slew limiter in the toolbox does not take a cv in, so good look unifying your polyphonic portamento, if...that's what you're after? 4 s&hs seems like overkill, so maybe that's why there are four tool boxes..unless they're for inverting the envelopes?
"it I am thinking of using to turn the lowest voice of 4 voice poliphony into a rythmic beating bass line"
this makes it sound like what you're after isn't like true keyboard polyphony so much as "multiple voices doing different things". if this is what you're after, me personally, i'd suggest some different sound sources or filters, etc.
i'm also separately confused by the proclivity toward mixers here. each of the four voices has a dual vca w/ 2x1 mixing, then the far right has another pair of 3x1 mixers and 4x1 mixers. i may be over thinking that one, and you may be intending these to serve also as cv attenuators. i guess it's possible you're intending on taking the already complex signal from the vcos and mixing them with some other output from the same vco, and having that serve as the complex voice. that'd be rad as hell, but at the same time, yield so crazily varying results between voices, that i'd just look into different types of voices.
overall, this looks rad as hell, and if you've got the $$ lying around, go for it. it just seems to me that it's got a ton of redundancy, depending on what you're after, either in terms of needlessly repeating the same signal chain (e.g. attempting polysynth stuff), or making the same blip bloops with the same modules but playing different lines (pads and leads and bass with the same voices) where you could be employing other interesting modules.
i enjoyed this more than i should have...thanks!