A classic case of "The More You Know, the More You Know You Don't Know" afflicts the creator. You're building a software product, or painting a large canvas display for a big-shot client, or producing a broadway musical.
These are medium-term undertakings -- more than 3 months, sometimes up to a year, and hence there's a very tangible time constraint. Because of that time constraint, there's no time to second-guess the methodology once it's set in stone. Before you've started, you've chosen the canvas that you're painting on, you've decided on your audience (hopefully!), you've chosen the venue of the performance, etc. Sure, the type of paints may change, the props may be rotated, the deployment tools may change, but the high-level architecture doesn't.
But while the architecture is static, the requirements of the project may change. This isn't a change in vision, but rather, a change in implementation. Maybe we need a secondary canvas, a small one, but one that doesn't permit the use of oil paints. Maybe the platform that the performers are singing on needs to be 3m and not 4m tall because of height constraints in the performing venue. Maybe that software library you're using doesn't provide the functionality you need, and swapping out for another requires a 3,000 line refactor (not a lot, but enough to be annoying).
All of these little annoyances need to be tackled directly by the creator. It's obviously wrong to say that the final presentation is what the creator wanted. But it's also obviously wrong to say that the morphed presentation takes away from the overall experience of the consumer. Still, the tension of having to ship a less-than-ideal product (in the creator's mind) manifests itself in a form of self-doubt.
As a creator, I think it's important that we never try to avoid that self doubt. Doing so would remove ourselves from all the little decisions that make up the final product. We become managers, and not producers, of the product -- dispassionate instructors. The best of the best never let the task of leadership distance them from the creation process. The best example would be that of a good movie director - they have to understand the impact of lighting, of managing their actors, of weather conditions, of post-production, and another million intricacies that can end up affecting the final product. Yet that manage it, and out comes magic (Peter Jackson and LOTR comes to mind).
Let's make it a point to seek that feeling -- not knowing how far we've come, and not knowing how far we have to go. That's the definition of adventure.