Pull requests vary in size . There is a lot of effort that needs to go into this. First they make relays in sync versus putting the burden on the clients . But they probably won’t by design

Also, git only connects to a single endpoint so ……that will be annoying as an end user to work around when outages occur