It's also important to remember who you're presenting to. If they're people who really understand the process, the constraints, and - most importantly - where they want the project to go, then the feedback is likely to be excellent. Indeed, the resulting conversations (and ideally, that's exactly what they are) can be some of the most creatively inspired bits of the process. Also, very productive, in that they are likely to focus on the removal of obstacles.
On the other hand, if you're presenting to people who really don't have a sense of the big picture (and who are often insecure about their position and judgement), you tend to anticipate their tendency to (a) fixate on small, rough bits and (b) issue VERY specific instructions about those elements (often, while ignoring the rest). This is the exact mechanism by which incompetent managers paralyze creative talent, reducing any project to a derivative effort that hews as close as legally possible to "whatever worked before." In this case, management IS the obstacle, and good luck getting them to remove themselves.
In these situations, you're no longer focused on the work. You're thinking about dodging bullets and managing up, hoping to find the best 'compromise' between a great creative solution, and the need to placate the fragile egos of people who probably don't belong in their jobs. The awesome thing about Pixar is that they really seem to get this, and invest considerably resources in creating and staffing the kind of environment where people CAN iterate freely.
But the question is, would daily iteration reduce the rate of bikeshed[0] arguments? I guess I could imagine how it would make it worse, but it seems more likely that it would make it better: the people who know nothing about the process (and don't care) won't be at the daily meetings, and the people that know just a bit about the process will be there and will learn all the details of the big project (i.e. the atomic power plant, to continue the original metaphor).
One thing that might help is to prioritize "end-to-end", happy-path work highest. Once you've got the bare-bones, end-to-end functionality working, go back and start filling things in.
However, the "just do it" idea of the post is great.