Exactly!
I argue Nix is long-term more important because, being and ever more becoming a universal tool to underpin sustainable software development, it has a much broader scope and potential impact. At the moment, for a large but particular audience of Linux enthusiasts, undoubtedly NixOS is much more important than Nix itself. See the 2023 community survey results for a vague impression (please check the complete results, there is a lot of room for interpretation):
So it’s not wrong to drive community growth through NixOS as the entry point, but it’s not the only way. (Possibly what we’re observing is a result of driving community growth through NixOS to begin with.) Looking at it from the inside, NixOS as an end-user product is much less well-maintained than any of its constituent parts. I strongly doubt NixOS is the driving force of the ecosystem.
In my opinion, it’s not at all about NixOS, but tooling to make software finally stop sucking – I’d even agree with your provocative claim that NixOS is but a byproduct of that endeavor.