I'm always impressed when I see people who have spent such a large amount of time on a single project. Me personally, I keep flitting from project to project, creating a prototype, or a proof-of-concept before I get interested in something else and pursue that. I think a lot of people fall into these categories, breadth first vs. depth first. It's probably best to seek a balance between the two, as then you will still devote enough time to polish the outcome, while still having the time to explore many completely different projects.
Yep, that's me, too. Always coming up with new ideas for great projects, moved by a great spark of curiosity, but usually those projects just tend to die because after some time I've learned most of what I wanted to (I start getting into the boring details, like polishing, bug fixing, etc) and also because new ideas and new stuff to learn just appear and take all my attention.
Even though I know that for every new project the chances of completion are close to 0, I still start them every time because it's still fun to experiment and useful to learn. And, who knows, maybe one of them might spark my curiosity long enough to make it a more serious project that might actually get to somewhere.
