But then it would probably take another month or more to polish it and fix a few annoying bugs. I'd want to integrate things like themes for when I feel like a change of style. And language servers for when I want more powerful integration, obviously. Oh, and integrated Git support is all the rage right now. But best make it modular in case my next client prefers Hg.
Then there would be a fortnight when I got obsessed with fuzzy string matching for search and replace and how it might interact with multiple selections to create a highly efficient quasi-macro system. That would inevitably lead down a rabbit hole of reading a dozen papers and watching several conference talks. OK, at least I might have got a nice speaking gig at the end of this one, at least if there were any local meet-ups actually happening by then.
But by that time it's mid-2021. I still haven't actually got a better editor overall than half a dozen choices I could have used the day I started. I also haven't made any money for the last three quarters, and the bills are coming due.
Just because I can fix something for myself, that doesn't necessarily mean I should or want to.