Most of the writings I've read on the topic of traumatic "recovered memories" have had an all-or-nothing attitude: Either they were inclined to trust all (or at least most) "recovered memories" of trauma, or they were inclined to distrust all of them. Until just the past few days, I don't recall ever reading any intermediate stances except by a few of the people who endorse truly wacky SRA claims. And I never encountered anyone who provided any useful, down-to-earth advice on how to distinguish between probably-true and probably-false recovered memories of trauma. Admittedly I haven't been staying up-to-date on the relevant academic literature, so it's quite possible I missed something important.
In my Talk To Action post "Exorcism and religious intolerance," Tue Feb 01, 2011, I refereed briefly to the SRA scare and "recovered memories." Someone with the name OldChaosoftheSun replied with a comment arguing that recovered memories are indeed a real phenomenon -- and so too did ArchaeoBob, whom I am inclined to respect, here and here.
In my reply to ArchaeoBob, I said I would be "interested to learn more about the circumstances under which a recovered memory is likely to be true." ArchaeoBob replied with what seems to me -- at first glance, at least -- to be a very credible, down-to-earth list of differences between probably-true and probably-false recovered memories. Of course, I'm no expert on memory and not really in a good position to evaluate the validity of his list, but at least it's not prima facie wacky in my opinion.