My lessons both come from my current side project (I will share it with you in a week or two, the website isn't finished yet):
1. Every project comes to the point where it hurts to continue. Keep pushing, the result is worth it.
2. You aren't as good as you thought you were when you started, but you'll be better than you ever were when you finish.
3. Sometimes, there's more points to a list than you'd expect.
4. One hour per day is easier than five hours a week.

How?
Well. I started out my project knowing some C#, but Jack shit about unity. I know most of what I might build will end up being shit I'm gonna regret, refactor and recycle later. But I don't give a fuck. Doing it is better than planning it.

It sometimes hurts to get rid of a carefully planned algorithm that took hours to build because it fails in practice. But it's the right thing to do.

Never plan too much. If I'd have planned this project out, I wouldn't even have started with what I'm good at: write code, break shit and experiment.

It's easier to progress slowly but steady. Look at some awesome games that have been worked on for ages while the public had their say (RimWorld, Project Zomboid, Dwarf Fortress...) as opposed to those that are developed behind closed doors and rushed to the market before Christmas or some other major event (Mafia 3, Fallout 76, Fallout 4 VR...). Progress slowly, deploy early, push often. And the one hour per day approach is a good way to do this.