If the average type of paper deviates over time, it doesn't matter whether the average number of citations is low - the type (direction) of research that gets cited shifts as well, and ultimately the direction of all research done shifts.