Glen...if you are talking about areas that periodically have periods of seismicity, and these are regular, that would be interesting to document. But, what you are saying...is that someone would look at an areas that is active on a random pattern, someone waits until it is not active, and then predicts for that area? If the time window was small and the predicted magnitude was large enough, this could still be significant (if evaluated by an independent expert by some statistical method). I agree that reading a prediction is more interesting if the basis is given.
Chris