Hi,
It depends on the definition of "idle". :) There is no universally accepted definition of an idle queue, it's just that years ago someone on the RabbitMQ
team decided that after 5 seconds of idleness (no new messages being published to the queue and no consumers receiving messages from the queue),
the queue is marked as idle and the Erlang process for this queue is hibernated to save some resources (you can learn more about this
5 seconds is a completely arbitrary value. For example, if you run "perf-test -ad false -f persistent -qa x-queue-version=2 -u cq -c 1 -P 4"
it's idle almost all the time (accepting the message and delivering it to the user takes a fraction of a second and for the rest of the time,
this queue process has nothing to do).

The same queue would be marked as idle often, if you published at a slightly lower frequency.
Quorum queues currently never hibernate. It's just a technical decision that can be revisited in the future. As with everything, it's a trade-off:
you save some resources while a process is idle, but you increase the latency when a message is published to a hibernated queue and you risk
a lot of additional work in some scenarios (eg. some users have large fan-out use cases: a message is delivered to many queues at the same time,
if all those queues are hibernated when this happens, there will be a sudden increase in resource usage when all of them are de-hibernated to accept the message).
If you are interested in the metric or hibernation of quorum queues, please share some details of your scenario and what you'd use this metric for.