[beermann]

This might be an unpopular opinion, but I've seen the exact opposite of this. Given too much freedom, I've seen designers and product managers derail entire development cycles. Compare a few days for a single designer against two weeks for even a small development team and not only have you lost an amazing amount of money, but the opportunity cost of not getting something into the users' hands and iterating on it is huge.

For the same reason I've also pushed for aesthetic changes to be moved to following iterations. If your wireframe is solid and agreed upon up front, then iteration can happen on both the development and the design side as you move forward. I think it's alright to release something that isn't fully baked because what you learn in the end is so much more valuable than trying to get it right the first time.

Bottom line: get something out the door and iterate. Too often people don't do either. Either they try to perfect it and take too long to learn from it, or they push it out and move on to the next thing. Nothing is going to be perfect the first time, so just get it out there and continue to work on it. Just because you're targeting a feature in a single iteration doesn't mean you should forget about it afterwards.