Though, this would require constant connection to a full node & my script currently only switches nodes when it encounters failures - I'd need to use a latency monitoring script to subscribe within acceptable latency ranges. Good idea though!