The biggest problem with GCP outages is that they've had multiple network issues that affected multiple north American regions at the same time. So maybe there

The biggest problem with GCP outages is that they've had multiple network issues that affected multiple north American regions at the same time. So maybe there are less of them on a per service basis compared the AWS, but when they do happen they are far more serious. The only AWS has had that came close was the us-east-1 s3 outage. But even that didn't break networking to running instances.