[Master/Worker/CMR - RFE] Update validation scripts to accept worker and master nodes on different models

We generally deploy Charmed Kubernetes with all-integrated k8s masters and workers. This means, for example, on a cloud network, master and worker nodes will be placed in the same network.

However, it makes sense to have different networks for masters and workers and free operators to enforce different firewall and routing rules to those networks.

One solution would be to use spaces and have different spaces per-application. But not all Juju providers offer space concept, such as vsphere or openstack.

An alternative is to create k8s masters and workers on different Juju models and relate them through CMR.

That allows operators to select which networks each node will come up and enforce different policies for each.

To achieve a complete Charmed Kubernetes deployment using master/worker CMR, however, validation scripts need to be adapted. They need to take into consideration that a second model only for workers may be present on this deployment and offer the corret options for that.