Seems to me that if there have been, say, 10,000 such cases and that 2,000 have turned out to be hoaxes then, logically speaking at least, there is a 20% chance that that has happened in the Allen-Farrow case, either by intent or inadvertently; as you may have noticed and as Robert Trivers and Richard Feynman have argued, people do have a tendency to fool themselves.