Separate, lower priority drift detection run queue

Instead of the current behavior of drift detection runs being skipped if there are not available workers, it would be nice for there to be a separate queue for drift detection runs. The separate queue would have to be prioritized lower than the planned/tracked runs so that it is never pulled from when there are planned/tracked runs waiting to execute.

This prevents users from having to space out drift detection runs evenly during off hours to ensure they all get run. It is both difficult to maintain and initially configure. There are probably many other solutions to the same problem.