Awesome! That's pretty much exactly what I'm looking for, though I'm interested to see how easy it is limit certain peers to certain functions. Not everyone has resources to crawl and index pages, but a lot of people can store the index.
I'm interested in having client-side web storage, so you can participate in the network by just having the search page open (opt-in of course).
I'm honestly not actively working on it, but if OP provides the database and/or crawler, I'll do some research on feasibility.