Oh going to get me banned yet, but there is actually only incidental evidence the stuff works on anything (meaning no real research or studies have been done and only reports from people that spent what ever amount it is sent a letter)
Second called into question those same letters. 2 in particular were scrutinized and found to be inaccurate at best and fraudulent at worst. (author from WWM denied sending in one with his name attached, NSF was supposed to remove it from site and I have no idea if they ever did)
A second had some embelishment via creative editing according to that sender.
Add to this NSF pays to display ads here and you can see where things get pretty sticky...