That one also appears in the llog posts if you follow the links: http://languagelog.ldc.upenn.edu/nll/?p=478

As far as I can tell, their critique is correct in that the research Brooks cites isn't enough to support the claims that he wants to make.