(This post was originally published on Parasite's "sorta-creative director's" blog here).
Game development is cool, we all know that (or at leas those working in the area know that). When working on Parasite we've been making a many iterations on how the levels are designed, so we keep a "consistent" experience (so to speak), meaning that the entire set of rules works across the levels.
This same iteration process is also being used to find ways to tell the story (while rewriting parts of it too, heh). This isn’t only being used to make a better story, but also to present it in a more “dramatic” way, so you uncover the different layers little by little (or at least I hope it ends up working that way).
So the downside is that you have to be ready to ditch some of the work you've already made, and this can be kinda hard, but these kind of choices help you make a better product.
I once heard that movies are not finished, they escape. I can say the same about games or any other medium, since it's finished when it's polished enough and ready for primetime, but I think you can never say "ok it's finished" since you can always find new things to improve and others you're never happy with.