It seems very dangerous to diagnose as such. Perhaps someone gets depressed every three months near end of quarter when their hours get cut?

Once someone demonstrably has issues with mania, yes, sure, I understand wanting to prevent it, but in a lot of cases I see very little separation from actual mania and "I felt good one day" (which if someone were depressed may feel very weird).

There's also strong arguments that one or two manic or hypomanic episodes are not enough evidence for a bipolar diagnosis.