Now create a C program that reproduces a person singing, with all the variables that a real singer can act upon
Because if you try the same on a big analog modular you won't get very far at all - less far than you can get with code, given that very basic singing synthesis was already a solved problem back in the 1960s, and systems like Vocaloid have been commercial products for more than a decade now.
https://www.youtube.com/watch?v=hDsJCJ02184
...Which is the core problem with analog synthesis, whether it's a modular or a box-of-knobs synth: it's the synthesis equivalent of writing code in BASIC.
You can do it, and it's good for certain kinds of jobs.
But even though a wall of knobs and wires looks like the last word in musical openness, it actually puts some very rigid limits on the complexity and sophistication of what you can do. And worse (IMO) it trains you to think inside those limits instead of encouraging you to explore beyond them.
To me, this statement implies that you either don't understand what a modular synthesizer system is or you haven't kept up with the times. I mean, you literally build your own system by choosing your own components and patching them together however you want. You're free at any time to pull the patch cables and start over. Add one new module changes the entire system.
Other than the fact that you need to be able to plug it into something, I can't imagine anything being much further from the definitions of rigid and limiting. It would be like calling Pure Data limiting because it's a visual programming language. I don't know, it just seems silly to me.
There's a ton of small outfits in the Eurorack space doing really great stuff. To any one who is curious I'd suggest checking out some manufacturers like Make Noise, Mutable Instruments, ADDAC System, Modcan, and there's just too many to name.
