EQF, that is jumping way ahead of where Shan's method is. According to Roger, there is little or no evidence from prediction that it works (Roger, correct me if I am wrong), and there simply is no logical evidence why it should work. It seems impossible that it could work, if we understand it correctly (and I understand it about the same way that you stated in your post). So, why think about applying it, is you are interested in doing good? Applying it would do harm.
Chris