discuss the scope of units that should be covered by capd CI jobs

To get the kubeadm-capd and rke2-capd CI jobs to run on generic gitlab runners, we had to disable the "monitoring" and the "workload-cluster*" units in 5640dd34.

We need to decide what kind of coverage we want from those jobs:

• increase the coverage to cover all components: is it sustainable in terms of gitlab runner resources in the future ?
• or, keep the coverage restricted to a smaller set of units