This is a good analogy, right down to Git's painful UX. Nix is powerful and difficult to understand. Often you'll find yourself in a spot where either you need to take hours or days to understand some fundamental component, or you can just paste a magic incantaion you found soewhere and move on. Good luck googling, it's almost as if they designed the Nix language to be unsearchable. When you do find an answer, it's often 3 years old and completely obsolete. Definitely recommend having a chat window open with other Nix users while working with it. When starting a meaningful project with Nix, make sure you're not on a deadline.
I like Nix, a lot. But I'd currently recommend it for a pretty narrow type of user/use case. Like Git.
I have to agree regarding the UX, though I would like to add that it is the language per se that is complex, but its standard lib/“nixpkgs” lib.
Also, not sure I can see where would you not recommend git? Is that a typo? It sure has a bad UX, but it is the lingua franca of version management either way, that even junior devs will have to fight to understand to become even remotely useful. And this might well be the case with nix as well.
I would not recommend git for organizations with monorepos, and not for companies that need to store large binaries (like games; git-lfs is a start but still a royal pain). Also I probably wouldn't recommend git for projects that are significantly more media than code. Depends on the team.
Git dominates open source, yes, but I wouldn't call it a lingua franca in general. Most companies I've consulted with were doing just fine with something else.
Time to start shilling it where you work. It has a pretty steep learning curve but it's worth it most of the time. In some industries (e.g. Haskell-adjacent industries) almost everyone uses Nix.
