advice that may or may not be relevant
@colinstu I generally recommend people to pick a project they want to build, turn it into a series of increasingly complex steps, and start with the absolute simplest version of it. I've done a lot of professional tutoring (heavily self-directed) and that's the only thing I've found to work reliably for people.
For example, the final goal may be building fedi software like Mastodon, but the absolute simplest version is "page that shows a bit of text", then it becomes "page that shows a bit of text *from a database*", and so on, until eventually some day you might get to full-fledged fedi software, having learned a lot along the way.
This doesn't help if you can't find a project you want to do, of course, but I've noticed that a lot of people (sometimes subconsciously) try to limit themselves to "simple enough" projects, when really it can work with a project of *any* complexity, as long as you treat it as a learning project you may never actually finish.
I don't know if that applies in your case; if not, then my suggestion is probably equivalent to what you were already considering, and can be safely ignored :)