I currently need (to know) it and would rather not.

Especially in a multi cluster, kustomize, istio, argocd, Jenkins on bare metal over multiple physical data centers.

To me with intermediate data dev ops knowledge this is quite a level of complexity.

True but frequently people claim that it is bad and nobody needs it. Whereas I would say Kubernetes is a bit like a bucket excavator, you probably don't need it and a shovel would do you just fine but that doesn't mean nobody needs bucket excavators.