Wow...hm, a lot of this needs to be a lot smaller. As in, your VCF and modulation source take up 42 hp in a 104 hp row, which means that two basic functions require over 1/3rd of that row. This makes no sense.
If you've got a Doepfer Monster Case, then yes, I could see justifying that sort of attitude toward panel space. But this is a build in a small cab, where you simply DO NOT have the luxury of spreading out like that if you're trying to make a very capable modular. Plus, some of this might not be the right sort of gear for what you want, which includes "Atmospheric sounds, experimentation"; the Fusion Modulator's trapezoid gens actually are fairly FAST for the sort of very slowly developing sounds that work as drifty, atmospheric bits. And with all of the tube circuitry...did you check your current draws? Do these work with a Mantis? While that cab has a 3A +12V rail, note that this is broken up into three 1A "zones", and those tubes are sucking up a lot of current.
And here's another "warning flag": "any you would recommend that aesthetically look nice". OK...go look at a saxophone for a hot minute. While we all know the shape and such, because it's easy to use as an iconic form...the network of keys, levers, pads, cams, shafts, etc etc all over this form isn't exactly very "aesthetically nice", no matter how the builder tries to nice it up. It still looks like some mechanical engineer's worst fever dream. But all of that IS what makes the sax work.
In short, don't worry about what the build LOOKS like. In all likelihood, that front panel will be smothered in multicolored patchcord wirebarf anyway. Focus on FUNCTION. Sure, an all-black modular might look snazzy...but when you record it, does the blackness of that front panel influence the sound? No? Then don't bother with that aspect. Optimize the sonic capabilities while, at the same time, reducing the cost. F'rinstance, the Fusion VCO2s...OK, fine, they're black, they have tubes. But they take up 28 hp between them, plus tubes in the VCO...ah...really, you want to introduce those nonlinearities further down the signal chain, and have rather precise waveforms at generation. And together, they're $672 for just two VCOs. But at the same time, you could drop $15 more and get three Pittsburgh Lifeforms Primarys...which are more capable in terms of waveform manipulation, modulation capabilities, etc. No, they're NOT black. No, they DON'T look like any of the other modules. But none of that really should matter; you're making a musical instrument, not a movie prop, and visual aspects really should NEVER be a prime concern in that process, especially if they force you into a poorer functionality choice. This is just one example here; there are others, if you start looking.
Honestly, I'd tear this down and chalk it up as a learning experience. Some useful things to remember with MG, btw...
1) Your initial builds will NEVER work. Nobody does one-shots and nails the result. Nobody. Creating a modular build you'll want to use and live with for years takes time and a significant refinement process; fast results only yield crap that we'll be seeing chopped up and put on Reverb/eBay in several months' time.
2) Don't define the case FIRST. Instead, start with a much larger cab. Either you'll...a: find that the case WASN'T "too big", or b: you'll be paring the build down to eventually arrive at your desired build size, with ample space in which to do that.
3) MG isn't a video game. There's no score here. It's not a timed exercise. Get used to the notion of exploring the site in deeper ways, studying modules, existing builds by knowledgeable users, and the concepts behind modular synthesis. Trying to rush this is a good recipe for wasting money.
4) Understand that a good result WILL involve certain amounts of technical compromises. You simply won't be able to cram in every single function you want, so it's important that you sort out how to arrive at those compromises within the "box" of how you work on your music.
5) The amount of f**ks given about what a build looks like should be precisely concomitant with the page count of that storied tome "Famous Antarctic Television Personalities of the 16th Century" (ie: zero). And...
6) The boring-looking stuff is what makes the sexy-looking stuff WORK. Add unlimited snazzy modules at your peril!