Based on recent events following updates to the servers and hosting, it's become clear that the User Performance Chart found in Reports | Auditing needs to be re-hashed in some way. Currently the only definite sources of latency are the server (in green) and the client (in blue). The "Network" indicator is so broadly defined by this report that it leaves a huge array of potential variables including (but not limited to) the host location network, ISP, infrastructure, and routing relays within the wider internet as a whole. This provides almost no diagnostic value unless the report shows green or blue bars, and even those (in the case of at least one clinic) aren't entirely reliable.