If January had the highest number of jobs since the Great Depression and we have returned to the March mean, this could still be true while there was absolutely nothing interesting going on.

If there was an unusually high peak of seasonal jobs just prior to the months in question, that would be an utterly meaningless statistic. If it's a moderate random drop coinciding with the end of a somewhat high seasonal number, it's still not relevant to long-term outlook. That's why monthly numbers are difficult to use in isolation.