Node pool resilience clarification needed

Documentation here: https://rancher.com/docs/rancher/v2.5/en/cluster-provisioning/rke-clusters/node-pools/ says:

The benefit of using a node pool is that if a node loses connectivity with the cluster, Rancher will automatically create another node to join the cluster to ensure that the count of the node pool is as expected.

However, if I terminate or stop the AWS node it is just marked unavailable in Rancher UI. I thought this was working at some point, but can’t be sure. I am running 2.0.7. Seems like a bug to me, but perhaps there is config setting or something to enable this?