I also understand that its UI limitations are also part of its strength, as being fundamentally text-based results in a consistent UX no matter what esoteric modes you activate.
I wonder if there's a middle ground; say, by allowing for frames & windows to have "chrome" that can take up some space on their edges and allow drawing widgets via GTK or Cairo and plumbing input events back into Lisp. Then we can have stuff like sidebars, status areas, code lenses, etc. that behave like modern editors, and people that don't like it can just leave the "chrome" option disabled.
What's stopping you from doing that in emacs now? That sounds like a bunch of stuff we can already have?
CodeLens is just some text properties and tool-tips/pop-ups.
I hope that McCLIM (https://common-lisp.net/project/mcclim/) can some day fill that role, or a project that is similar to McCLIM.
I cannot imagine a better way of interacting with a computer than to have objects that can be manipulated both interactively and with Lisp code. Emacs is already close to that ideal, but I think we could do even better.
I managed to accomplish this somewhat with Lua. Designing the engine so that all the cool things like creating and equipping items, generating maps, etc. are in public APIs you can use interactively has given me a lot of power. I'm not sure if the benefits offset the maintainability tradeoff, but I'm somehow continually excited to sit down and mess around with it. Taken as a single-user experience, being the only person to understand the system and write more code, it is almost a dream come true.
I wonder if I would feel differently about this when playing a sci-fi RPG... hmm...
I actually kick-started my maps with rogel-ike. An Emacs roguelike game. Check it out.