I often find that the likelihood of my finishing a project is tied to how 'complete' the idea is when it first comes to me.
People wiser than me have argued that it's a bad idea to develop a game starting from a story, and other such people have made the argument that the opposite is just as inadvisable. At its most basic level, I think that plot is very amenable to games - I mean, what is this whole characters struggling to reach their urgent goals thing if not a recipe for gameplay?
Obviously there are bits that aren't so clear, like how to tell a story where the player-controlled characters don't win all the time without making the player feel like they should have been able to avoid defeat. And I think that's interesting to consider in light of a lot of indie/amateur games, which adopt a more short story/flash fiction style, which skips past that part of the plot anyway.
But I'm getting away from my point, which is that if you can come up with your characters and their struggle all in one go, it's better than coming up with one or the other separately, and when you're battling with one incomplete idea and a whole-formed one pops into your head, you have to go with the flow. I posted a little about a high-seas pirate game I was working on a while back. I still intend to finish that, once I figure out exactly what it is. In the meantime, I've been working on something completely different. You'll hear about it when I'm too far along to back out.
As for where Sky Spiders figures in all this, well, I guess that's the exception that proves my point. The point I stab myself with every Sunday.