PreStage incorrectly reports batch node readiness if quota has been reached

[jormdav]

We'd a situation in our batch account where we attempted to provision more nodes than our 500 vCPU quota allowed:

However, when prestage was waiting for all of them to idle, it incorrectly reported that all 350 nodes were ready, when only 349 of them were.

The above example was created for illustrative purposes, where it bit us was a separate pool of 1 node that couldn't be created, but passed its prestage logic and moved onto enqueue/run, whereupon the batch job eventually timed out: