Comment by KingMob

> I'm sure people already do. But you can't truly test things at scale.

Do they, actually? I got the impression from the article that little revocation testing of any scale happens. If so, then even a limited test would be useful.