I don't know if that should surprise us. AWS hosted their status page in S3 so it couldn't even reflect its own outage properly ~5 years ago. https://www.theregister.com/2017/03/01/aws_s3_outage/