I’m assuming GitHub has a fair amount of database/cache overhead for most things, especially branches. I think that most things the web client sees are all database content and that there’s no usage of git/filesystem in any hot paths for web views.

So I can easily see why having many branches is more storage than the same number of commits.

It might be something silly like the number of items in the Branches dropbox menu.

We appeared on a list of the top 10 most egregious users of github so I assume they had database entries for these…