The link doesn't support your claim. There is no evidence of a reversal, and the original researcher most subsequent authors cited apparently never suggested there was, only that it was inconsistently applied.