Add worker pool telemetry

We already have queue telemetry, but it would be cool to monitor the health of the worker pool too.

I would love be able to toggle between queue and worker view.

The [psutil](https://github.com/giampaolo/psutil) package allows us to telemetry per node. Via Redis PUB/SUB, we can feed it back to the inspector and display it. I'm thinking of a similar interface like btop or htop (two top alternatives with a nice TUI and graphs).

Some metrics I would be interested in:

node mem
node cpu
node task throughput (tasks per minute)

worker throughput
worker memory allocation
worker CPU allocation

The graphs for throughput, cpu & mem should remain at the same position while displaying different data depending on what is selected.

The selection should be happening via a tree. Queue -> node -> worker
While nodes can repeat per queue, if multiple nodes are listening to the same queue.