I wonder if, instead of changing the core *behaviour* by adding the
option to look up the nodes using hostnames, we shouldn't improve the
CLI so that it's easier to *extract* information from the returned
responses. This is based on the idea that most of the pain we feel when
using the CLI comes from the awkwardness of having to deal with
JSON-formatted output. While this is fine when using the API in Python,
JSON is clumsy to use on the command-line (even though I know that they
are workarounds, for instance by piping the JSON blob into a Python
command).