Having to reset or recalibrate other old systems that were disrupted by newer ones going offline makes sense to me. If servers were providing Network Time Protocol and older clients drifted without it, that could cause them to be unable to rejoin a domain. I’m speculating wildly, but it’s an example of how losing important infra can cause issues even after it’s restored.

It didn’t say “non-windows” it said “served by other providers like IBM”. It could easily be Windows servers in IBM’s cloud and wouldn’t ya’ know…IBM uses Crowdstrike.