The test currently depends on the overloaded fsm workload's teardown check to implicitly detect duplicate measurements (it checks for disparities between the sharded collection and the unsharded collection).

This would add an explicit check for duplicate measurements as part of the teardown function for the new workload.