If anything, I think your setup would be bigger if you had all these modules, because the real juice comes from modulating and manipulating voices and sounds with control voltage and routing the signals creatively more than methodically layering and combining lots of voices (which is kind of better left to non-Eurorack hardware). Think of a jazz trio - a bass, saxophone, and drums. Despite only two mostly monophonic melodic voices (with some opportunities to do chords or whatever) and and a few drum sounds, the possibilities are immense because of how much they can manipulate timbre with their playing. Albert Ayler's Spiritual Unity is a pretty good way of demonstrating what I mean - that music is massive despite being basic instruments available in reasonably well-stocked high schools. That level of flexibility per voice is what makes modular special relative to other formats despite essentially making the same sounds as any other interface with the same circuits inside of it, but it comes from modulation and utilities. LFOs, envelopes, VCAs, mults, CV mixers, clock dividers, sample and hold, switches, etc. - the boring stuff that was pretty much already figured out by the 1970s.
For example, the small system you made has two sound sources (three if I'm being picky, but I won't count the Tukra). One does include its own support modules, but that will just get you to the point of making regular synth sounds. When you consider that the Tukra only sequences gates and you still have to get notes into the thing, you'd probably do well to stick to a general "one sound source per row" rule and go from there. Theoretically, if you said "I want to get a setup with at least two synth sources and some unique sequencing/drum options", I would recommend a Syntakt or Analog Rytm before this modular setup, and you'd make the same sounds better and more conveniently.
Instead, my advice for your first modular setup is to go the other way and just think "what is the most badass single voice I can come up with?", and maybe add that onto the Tukra for a cool start. You can even use the Pam's to quantize modulation from weird modules and create generative melodies, so maybe keep Pam's if you're going to start with a pure gate sequencer like this - it will help cover for it and integrate it into a less rigid framework well. After you have a full, tricked out single voice figured out, the setup ideas will just be flowing and it will be easier to make it more expansive and nebulous.
Also, if you need 104 HP for your first rack, go for it. I just kinda said 84, but a Mantis case could be wise if you know you want to do this stuff.