A pity this science news article does not include a link to the study, as the results sound interesting.
I wonder how a super-recognizer might have fared as a study subject in this study? Would supers have the self-confidence to avoid the impairment in performance that comes with anxiety, while not being so over-confident that they fail to take the extra care to see through racial bias while memorizing and recalling foreign faces? Are metacognitive skills an element of super-recognition? Could this be a clue to why face memory is a cognitive ability that peaks in performance at a remarkably late age in human development? (Please remember, readers in academia, that if you use without acknowledgement original ideas that you have read at this blog, I will not be pleased at all.)
Payne, Rob Anxiety increases error, but not bias, in facial recognition. Science Network. November 20th 2015.