More explicitly state that a node does not have to be failed over to be removed via the CLI

I've seen a few support cases recently where customers are under the impression that they need to failover a node before removing it from the cluster and rebalancing.

Whilst this is documented in detail for the Web UI and not true for the REST API, for the CLI this is only inferred, and not explicitly stated in either the rebalance command reference or the "Remove a node and Rebalance" page.