Consider relaxing alerting rules for urgent-authorized-projects

This shard is deliberately constrained due to the large number of jobs this worker can start. This means queueing and execution SLI violations occur quite frequently, paging the EOC in the process. Since the concurrency constraints are by design, perhaps we should look into how we can tune the alerts so that the EOC is paged only when there is an actual problem with this shard/worker, rather than every time someone does something to trigger this worker.

This ticket was created from INC-2647 and was automatically exported by incident.io 🔥