This is exactly what I do, get the fun done and leave the boring stuff for later. The trouble I've found with this is that often, the boring stuff is also the important stuff. Example, I normally get the basic game world up and running then attempt to put in control and GUI elements later. Not only does this leave a big pile of steaming bore, it also meant (for me) that the boring stuff was even more difficult to do, having given it no previous thought when implementing the fun stuff.

I think a lot of games that grow out of fun little prototypes have this problem. I think it comes down to people wanting to expand prototypes into production. What you should be doing is proving your concept with the prototype, then once you decide to go forward with a full game do a little bit of planning, throw the prototype away, and start over with a solid plan for a completed game. I think the planning phase is something a lot of people miss when they work on personal projects that would save a lot of stagnation and pain.

I make a handwritten list of the hundred crappy things I have to do:

• Make a settings screen
  
• Make a music volume slider
  
• Make a SFX volume slider
  
• Make a worldwide high score list screen
  
• Show your worldwide rank on your own high score list
  
• Make ongoing sound effects stop when switching to the menu screen
  
  And on and on and on. There's a real satisfying feeling to physically scratching something off a list. It's sort of like doing chores--the list seems interminable at first, but if you just knuckle down and do it, it's often over surprisingly quickly.

Iterations are good, keep it up.

I'd also assign certain days of the week that you have to work on it, regardless of what happens. Lately Saturdays have been working well for me, because I also have a full-time job. On Saturdays, you have to do your project, no matter what happens. Earthquake, tsunami, you gotta stick to your desk. Just kidding :-), but you get the idea. This forces you to keep working on it because you have made the commitment. Usually, the motivation picks up again after you have worked on it for about 30 minutes - 1 hour.

Don't work on it on other days, no matter how tempting. Save that excitement for later. It also allows you to rethink your ideas and approaches, and many times I came up with better ideas during this time. As soon as Saturday hits, it's just a straight coding session, resulting in better and polished products.

I'm not sure how many other people would agree with this, but one method I use is to keep a bunch of projects going at once, so that when one starts to feel a little stale, I can immediately jump to another without losing any net productivity. The pitfall with this is that you risk never getting anything done, though, so you have to be careful about it.