The benchmark is on web requests per second, so it really needs to cite each server's location or ping latency relative to the machine doing the benchmarks. It's unfair if one of the servers is on the opposite coast of the other 2.