The magical belief that p2p is always superior to server-client for this kind of application has to be refuted. I live in Kingston Ontario, a medium-sized city between Montreal and Toronto. There are essentially two high-speed ISPs: Bell, using fiber, and Cogeco, using cable. Latency between peers on the same network is usable, but packets from one network to the other, even a few blocks apart, are routed via, for example, Toronto to Chicago to New York to Montreal and finally back to Kingston. It's not the distance travelled that was problematical; it's the latency introduced at each intermediate node. By setting up a server at AWS in Montreal, the latency for peers on both networks became usable.

Also, p2p requires O(n^2) links whereas client-server has O(n) links, so if the number of peers increases, the total travel time difference will be more significant.