My biggest problem is in "no need to get fancy". Do not know how to overcome that urge to overengineer everything in the start and how to be satisfied with minimal working thing even if it looks like crap.
For me, I use three other feelings to combat my high standards.
One is fear. In particular, fear that I'm building a bunch of stuff that won't work. It's the same fear that drives me to write automated tests, or at least to manually run something and see that it works. Except here the fear is that it might work for me but not for my audience.
Another is rigor. This part feels very sciencey to me. I'm trying to test the hypotheses that are implicit in the thing I'm making (and that I've hopefully at least tried to make explicit before I start coding). I want to test my hypotheses as early as possible, and when I don't I start to feel like I'm not being rigorous.
The third is joy in results. The early tests are only looking for one step forward. If that's all I'm looking for, it is really exciting to see that. And so I temporarily set aside my concerns about all the things I'm not testing for and celebrate the wins when I get 'em.
When you don't have an hour a day to work on your side project, but 15 minutes a week, then you do what is needed for it to start working, even if it's not perfect or far from what you wanted.
If you have plenty of time try to make the smallest thing that does approximately what you want. Then iterate over it, but always small but fully functional incremental changes. Don't be afraid of having to throw it away, no need to get it right the first time.