This is what happens when performance reviews are tied to outages. I spent over five years at AWS, and even internally we treated the AWS status page as a running joke -- a bunch of us used an extension which just set the level of every outage to one higher than had been posted (i.e., green i/blue diamond was yellow, yellow was red).

The problem is, organisational leaders have performance reviews (and bonuses) tied to service health, and they'll avoid declaring outages to avoid looking bad.