I find this as well. My motivation for new projects is 100x compared to my motivation to finish existing ones, and I feel part of it at least is because on the existing ones I've reached a point where I'm not learning anything from finishing them. The flipside is that I have a lot less work that I can point to having done, compared to how much I've actually done. Having 5 completed things to show is better (for "achievement" purposes) than 20 projects at 80%, but the rewards as you say favour the latter.
This year I'm thinking about only working on projects where I'd actually use the outcome for something. If it's a project for the sake of learning library X, I won't do it. If it's a project to solve problem Y for myself, I'll try and work on that, learning anything I need along the way.
This year I'm thinking about only working on projects where I'd actually use the outcome for something. If it's a project for the sake of learning library X, I won't do it. If it's a project to solve problem Y for myself, I'll try and work on that, learning anything I need along the way.