For fun, I took all the major projects I've completed, and split them into two columns. In the left-hand one, I wrote "where it started"; in the right-hand one, I wrote "where it went". (The joke should be obvious.) The right-hand column just contained an image of the finished book's cover. But the left-hand column contained the earliest recognizable version of the idea, which often had so little to do with the finished product I might as well have not bothered listing it. Was that a problem? Not really; in fact, it's a good thing.
Most projects don't end up anywhere near where they start out, and for good reasons. That first spark that kicks everything off is often just that, the first spark. The more projects you kick off and complete, the more you see where you started doesn't have to be where you also end up. You may find that original spark to be a great motivator, but it may open the door to a much larger set of inspirations that in time lead you quite far from the original. By no means are you bound to slavishly include the original, especially if better things beckon.
One example of this is a project already in progress (sort of): Charisma. I detailed the wonderful and weird genesis of that project here. It started with a piece of clip art, turned into an in-joke between friends, and then expanded from that into something far more ambitious. Where it eventually ends up, I hope, is in an even better place than all of those things put together.
Anime and manga often have this trope wherein some unkillable monster has this one vulnerable spot somewhere, a "core", which must be destroyed. This seed idea often feels like that core: you're worried that if you let it drop out of the picture entirely, the underlying reason for the work will also vanish. Thing is, it's not even the same work anymore. The original idea may well be nothing but the raft you ditch once you've reached the other shore.