Git is perfectly capable of having a repo that is considered the "single source of truth". All you have to do is tell everyone which repository to consider the "main" one.
The reason it's fine to use any random developer's git repository to restore is because the data's integrity is guaranteed by git's design. Let's say you had some disaster and lost your repository. Suppose you found a random developer (that you don't necessarily trust) who has a git repository that is a clone of your old, destroyed repository. You can safely and confidently restore from his copy so long as they have a commit id (hash) that matches exactly with a known hash from the original repository. If the hashes match then his copy at that commit is guaranteed to exactly match, bit for bit, your original repository. Compromising that integrity would be extraordinarily hard.
Of course you shouldn't rely on random developers having copies, just out of pure luck. Make your own backups and keep them safe.