I hadn't seen that site before. I think it's really cute how they update their past predictions [1] after the fact [2] to improve their apparent accuracy, but you'd think they would have the sense to stop short of having "predicted" the Chelyabinsk meteor explosion last year.