jojy - 12:04 am on Jun 22, 2013 (gmt 0)
Essentially, what you meant by P2P search engine is to rely on user computers rather than on a central repository if I am not wrong. You will still need a central repository to store all your crawling information or if you want to categorize information and distribute it to peers then it may help serving faster queries geographically.
Personally, I think P2P search engine is useful if you have central repository stored in multiple servers (like google), you have cached common search terms and the cached result is distributed in peers.
I hope it will help you!