In terms of development time, that final polish pass (getting an area from 80% to 99%) takes almost as long as getting it to that 80% in the first place. Our attention to detail makes some cool looking stuff, but also slows us down significantly. 233 is a good example of this. One of the reasons for our recent shift in priorities is that we feel it's far better to first have the entire game at 80%, rather than only a small part of it at 99%.
In our eyes, we'll never get the game 100% - there will always be something to improve on, a tweak or improvement to be made. Releasing the game within our lifetimes requires us to prioritise which areas need the most work, so it's best we get the whole game to 80% before we start being perfectionists again.
I hope that makes sense.