Hi

The Cluster Wide Resources are still considered to be somewhat experimental.
You can find more abou their history here: https://www.securecodebox.io/docs/architecture/architecture_decisions/adr_0012

In general they are supposed to work exacltly like the "normal" namespaced scantypes but allow you to define them across the cluster for scans in all namespaces to use. Scans are still namespaced, so the pods will be created in the namespace of the scan.

This was in part designed to make the auto discovery easier to use by allowing you to thefine these ScanTypes, ParseDefinitions and ScanCompletionHooks once per cluster.