My early projects oscillated pretty widely in difficulty when I started. I'd take on something ridiculously difficult (but I didn't know it was ridiculously difficult at the time), and naturally fail at it. Then I'd take the skills from that and do something I knew I could finish. Eventually this converges, and the highs become less ambitious while the lows become more ambitious. I've found that I finish a significantly higher proportion of projects now than I did in college, and their difficulty levels more closely approximate a straight line than a sawtooth wave.
A lot of people are going to have trouble admitting this, but a diligent person of average intellect is about a hundred times more likely to become successful than a lazy person with a titanic intellect.
I can provide examples if necessary.
I know a lawyer of 16 years with a 145 IQ who is still making 55K a year. Ambition and diligence go hand in hand. She got the job right out of college and grew complacent, and now rarely attempts to find a better position somewhere else. Starting her own firm is almost certainly not going to happen. Other than diligence, lack of courage is also an issue there. Quitting your job and trying to build a clientele is scary.
I also know a very bright lady who cleans motels because she can't deal with people and just doesn't like working that hard. She works a few hours a day, at three separate places, with long breaks in between. She probably knows more about computers than I do; she's been manipulating them since before I was born, but she hates pressure, she hates deadlines, and she lacks confidence.
Trying to list all the unpublished author-hopefuls I know who just stopped submitting stuff would be an exercise in futility.
> I know who just stopped submitting stuff would be
> an exercise in futility.
That seems a little ironic in the context of this thread. :D
So, recently, I've tried to change my mindset to what you describe and build something small, useful to me, and that I can finish fast enough that I don't have time to lose interest. And I have a few other bite-sized ideas to do once I'm done with that. (my current tiny project: http://logmeoutthx.com/ I'm not so much interested in the thing itself as in completing it)
Important things:
* start with small things and set a specific completion date/feature spec.
* stop at a point where you know exactly what you need to do next. It's much easier to come back when you do this.
* use org-mode to plan/track your project. I like checking things off. The feeling of meeting mini-goals will keep you going.
Go forth and finish :)
EDIT: Your IQ is not relevant to your ability to complete projects. It is possible your feelings about your high intelligence is a problem in that it keeps you from being willing to experience failure. These fears should dissipate as you incrementally do complete projects.