Thanks for the feedback. Is there a programmatic way to get the IPs of all the current p2pool nodes? I could use that to have the site automatically update.
That is a good question for windpath.
It's really pretty straight forward, however you do want to consider the load you put on the network (i.e. how often you scan for nodes)...
Just pick a node, get the /peer_addresses endpoint, and crawl outward from there.
Example:
http://p2pool.org:9332/peer_addresses