Basically every screenshot of the "lost" TUIs look like a normal emacs/vim session for anyone who has learned about splits and :term (guess which god I believe in?). And people still use those near constantly. Hell, my workflow is generally a mix between vim and vscode depending upon what machine and operation I am working on. And that is a very normal workflow.

I use emacs, and kind of had the same gut reaction, but they do address it and have a valid point in that the IDEs they're talking about are "out of box" set up and require little learning to use in that mode.

Like, you can use emacs and I'm sure vim as an IDE, but what you have is more a toolkit of parts for putting together your own IDE. That can be really nice, more flexible, but it's also true that it isn't an off-the-shelf, low-effort-to-pick-up solution.

Emacs had some "premade IDE" project I recall that I tried and wasn't that enthusiastic about.

I don't know vim enough to know what all the parts are. Nerdtree for file browsing? I dunno.

With emacs, I use magit as a git frontend, a compilation buffer to jump to errors, projectile to know the project build command and auto-identify the build system used for a given project and search through project files, dired to browse the files, etags and some language server -- think things have changed recently, but I haven't been coding recently -- to jump around the codebase. I have color syntax highlighting set up. I use .dir-locals.el to store per-project settings like that build command used by projectile. The gdb frontend to traverse code associated with lines in a stack trace on a running program. TRAMP to edit files on remote machines.

But that stuff isn't generally set up or obvious out of box. It takes time to learn.

EDIT: The "premade IDE" I was thinking of for emacs is eide:

https://software.hjuvi.fr.eu.org/eide/