We currently convey cloud creds issues in ValidOIDCConfiguration and ValidAWSIdentityProvider conditions.
to decide if forcefully deletion should be applied and so potentially intentionally leaving resources behind in cloud. (E.g. use case: oidc creds where broken out of band).
The CPO relies on those to wait for deletion of guest cluster resources https://github.com/openshift/hypershift/blob/8596f7f131169a19c6a67dc6ce078c50467de648/control-plane-operator/controllers/hostedcontrolplane/hostedcontrolplane_controller.go#L284-L299
When any of the cases above results in the "move kube deletion forward skipping cloud resource deletion" path we should send a metric so consumers / SREs have a sense and can use it to notify customers in conjunction with https://issues.redhat.com/browse/SDA-8613