[API] expose each client's allocation percentage

It would be incredible useful to include how "subscribed" a node is on CPU/Memory/Disk

e.g. this is not how busy the cpu is right now, or how much disk space is currently used, but the amount of resources "locked" by allocations in the jobs.

this will make it infinitely simpler to writing auto-scaling logic since we could work directly on the nomad state, which is used to place new allocations by the scheduler, rather than trying to compute a similar metric out of band.

If it's in percentage, a float or just absolute numbers are all fine, as long as its something thats stable and ~easy to convert into a reserved resources factor for each node :)

No files, tests, or entry points are named in the issue. Start by locating the API and allocation data involved, then clarify the representation and scope for CPU, memory, and disk reservation values; done means clients can obtain a stable per-node allocation metric suitable for autoscaling.

Written by the indexing model from the issue text.