There’s no particular reason why VRR monitors couldn’t support the same frequency ranges as gsync. I assume that’s what the “gsync approved” monitors do.
As for being superior technologies, they’re basically the same thing. Gsync is in a separate module, and Nvidia mandates tight specs for validation like that refresh range, but otherwise identical.
Gsync also keeps a full FB in memory, but it’s never been entirely clear why that’s advantageous. Plain old VRR works fantastic on AMD GPUs, and I don’t see why it wouldn’t do the same on Nvidia ones.