I have the opposite problem. I play and play and experiment but rarely get to doing real stuff. I try to go very deep with my understanding before doing anything serious and through this usually discover how actually complex whatever I study is. And this complexity throws me off. The tool that I am studying becomes a subject of study itself.
This isn't an overfocus or overstudy problem, it is just a fancy way of getting distracted or procrastinating. You can learn at least as much by diving in to make an 80% solution and then polishing it up - as you can by deferring the work for more study. Maybe more, because there are some things you just don't learn until you actually have something and until you actually finish something.
I agree completely. When I find myself in such a situation I usually try to slap myself mentally and go on and actually do something. And from experience I know that I will actually learn more through doing. There is also matter of interning some things through actually working with them. That's why it's important to actually solve differential equations when you are studying them and not just read the theory. And yes, I am a procrastination champion alright.