I'm getting more and more ideas for stuff that works well with Leo. It continues to be great for the Scrivener project I adapted, and I think it has some promise for restarting Drawabox, but I don't see myself using it for any programming projects, at least not directly. It might work as a replacement for my use of todo.txt in some projects, I'm not sure.
It's a combination of, it's really great at representing multiple views of how a concept or task is relevant, a little iffy at manipulating those representations, and when I read the documentation for anything relating to code, I get hopelessly confused until I try it myself, which means I first need to figure out what to try and how to try it. It kind of feels like Leo is pushing a bunch of paradigms in implementation that I'm unfamiliar with, so I can't really judge their worth.
I still think it would be nice to get Leo to a point where it could substitute for my current usage of Sublime Text, or more of my former usage of Scrivener, but I don't know yet all the things that would need to change, and which of those are configuration changes, and which are code changes.
(For example, how do I make it change autocompletion behavior based on the active language? I don't want python completions in HTML text.)
Anyway, I'll be looking into that as I feel like it, but it's at least proving useful for now.