So, I've got good news and bad news.

The good news is, I've realized that we can use the idea of a merkle tree to make the maximum sector size essentially unlimited, which saves a lot of gymnastics for larger files, and makes file proofs a lot cheaper in terms of bandwidth.  It also makes changing parts of a sector reasonably priced without needing to download the whole sector. (before, you needed to download the entire sector before uploading any changes. Now, you just need to download log(n) hashes).

The bad news is that I've got to redesign and implement a few things to make it possible. But it makes file proofs more bandwidth efficient by a factor of 200, which is a pretty big deal because that was a bottleneck before and makes the minimum requirement for being a participant much, much less. I'm still hopeful about July 1st.