I think the difference for me, with Emacs (which is what I prefer and use), is that similar to linux, it does not go through thorough review and QA processes before release, and a lot of the features are provided by different people collaborating distantly in the open-source, on mailing list, with no dedicated team, and that has the benefit of having a lot of innovation, but also it means often things break, and it doesn’t always work in all setup, etc.
But, similar to Linux, and this is where I say Emacs is not as “easy”, as a user, you can learn to self-diagnose and fix all such issues yourself. This is the beauty of Emacs, but doing so is next-level compared to just someone who needs an editor, you have to learn to program in ELisp, understand the design of the Emacs code-base and how things are modeled, etc.
As a programmer, this appeals to me a lot, since I can program and know Clojure, learning ELisp is not difficult, and that gives a lot of power to customize everything and fix issues I face myself. And customizing Emacs and fixing issues with it is done using interactive REPL driven development, best of all worlds.
But I reckon if you are new at programming, are not a programmer, don’t know Clojure or Elisp or any Lisp, are not a fast learner, or don’t have time to learn, because of that I say Emacs is “harder”, because in my experience, it won’t work “out of the box”, and when it doesn’t, “fixing it” often requires programming like diagnostic on your part.
Similarly, I’ve find VSCode works more often out of the box, Microsoft controls most of it, and keep things more in check so they work, does extensive QA, and all that. You can still have extensions that don’t work, but you’ll use much less of them overall, and they can’t do as much as they are more restricted.
The actual UX, well I think I’d agree, if you come at it blind with no prior expectations of what “standard editing shortcuts are” or how a “GUI” should work, or what a “window” is compared to a “frame”, etc.