I don't know if that's the whole story? Sure, GitHub has more data to process (and way more compute), but even when switching between tabs within one repo (going from code to pull requests, for example) GitHub is still much slower even though its query shouldn't be affected by the number of repositories it has.