In order to make in-cluster Monitoring optional we should add an OCP capability.

Other then the API addition described in the enhancement above this entails annotating some of the CMO manifests in https://github.com/openshift/cluster-monitoring-operator/tree/main/manifests. We will retain a minimal monitoring stack to send telemetry, so the operator, Prometheus CRD and PrometheusRule CRD will need to remain (plus some supporting manifests). Aleternatively we cloud also generate two versions of the manifests.

See this thread for the order in which this is supposed to happen.