AFAIK it was a signature update, and the existing code couldn't understand the new signatures with a 2022 timestamp. IMHO, signatures still should have been deployed to an early ring before broad release.
These kinds of things are always tricky, and always easier to see in hindsight. Everything is a trade-off.
I can easily picture a scenario where there is pressure to reduce time to release signature updates, and someone looked at a process where there was early testing (that would have caught this) and said "Over the past 5 years, we've deployed hundreds of signatures with exactly zero failures; why are we still doing this process that adds a pointless delay?"
Is there actually an update here, or just some code that was probably sitting around for years until 2022 rolled around?