First, they mean doing a lot of work ahead of time costing things out.
Second, they are not an efficient use of resources. The value *I* place on certain bonuses may be quite different from the value that the community of supporters place on them. By letting the supporters decide, the value is maximized.
Third, it's an experiment I want to try. In the long run, it'll be more work (running the voting on what gets done, for example), but I think the value of what I'll learn from doing it makes it worth it.
I added the Challenge Coin as a way to thread the needle, and enable a lower tier.
Remember: my goal here is not just to do BGC, but to explore a new model for releasing.