Armada should surface the real problem when a necessary queue doesn't exist

[sync-by-unito]

Describe the bug

Armada returned this error message from a grpc client call:
rpc error: code = Unknown desc = [x-request-id: "efPZWLuVxvEjTom2UtgR3R"] job 01gv4gbqervamwfrjzvq05bwrt unschedulable: failed to schedule onto legacy scheduler because 0-th job unschedulable: no executor clusters available

But the real issue was that a queue doesn't exist.

Files Or Markdown that can reproduce the issue

Expected behavior
Armada should return an error message describing the actual problem. In this case the requested queue didn't exist.

Actual behavior

Armada returns an unhelpful error message.

Additional context

┆Issue is synchronized with this Jira Task by Unito