This build has got some very real problems, actually. For starters, I don't see any VCAs aside of what's in the Tesseract mixer and the M32. The mixer's VCAs might take care of audio levels...but how do you plan on using modulation to modify CV/mod levels? Basically, you can't. And part of that, also, is the drastic lack of modulation sources. Aside of the M32, I see only one very lonely Make Noise Function. No other LFOs, no ADSR (and variants), zilch.
Then there's the M32 itself. Mother 32s come with a case that has power already. And you've already paid for that case. So why pay twice to house the M32 and power it? It might be convenient, but economically, it sucks. Try this: take the price of your case + power supply, and then divide that by the total HP in the case. Then multiply that x60. Not so convenient now, is it? Even if you just count the one uZeus on that row in the equation, it's still a lousy bargain. And if you want (need) to add more (as noted), you'll now have to get yet another cab in which to put the various VCAs, LFOs, EGs, and other oh-so-boring BUT ESSENTIAL modules that this build just ain't got. So...take the M32 out, as you're going to need that extra 60hp real soon now for things that don't have their own cases and power, which is what should be in a Eurorack case anyway.
Next, things that are there that don't need to be. Let's start with the Takaab buffered mults. Again, aside of the M32, you have two VCOs. You don't need buffering to correct CV sag with just two VCOs. This problem really starts at 4 or 5 and above, depending on the frontend buffering on the VCOs in question. Now, I can see the mistaken idea of putting one by the VCOs...but the second one is for...what, exactly? Fact is, in a small build (which this is), you should avoid adding anything that you can replicate outside of the cab, and multiples are one of the big offenders here. It would make more sense to use inline mults or stackcables, given the space constraints...yes, even to replace the buffered mults here. Trust me, you won't see any detuning. Then there's the GSMN tube distorter; did you happen to read this in the MG listing: "PLEASE NOTE: As with all vacuum tube based designs, this module is very sensitive to noisy power supplies."? If this is the case, why is this next to the uZeus on the bottom row? And this gets into the whole concept of workflow/signal flow itself...and this is sort of a mess in that aspect.
Go and study classic, prebuilt modulars and semi-modulars...instruments such as the Moog 55, ARP 2600, EML 101/200, VCS3, and so on are the way they are because their workflow patterns are ones which function to assist the user. This current build doesn't seem helpful at all, which you'll find extremely annoying when your career starts to move and you're gigging live in a low-light situation and suddenly you're having a very hard time figuring out where you are on that scrambly patchpanel. Building a modular synth is just like building any other musical instrument -- it has to make sense from a musical standpoint, and it needs to be carefully thought-out before throwing money at it, because much of that money will wind up going up in smoke if you've proceeded with that process before A LOT of careful preplanning. That Neutron may have been inspiring, but it also may have given you a lot of wrong ideas, because they're definitely showing up in this build.