A shortcoming of this article is, imho, that it is taking manufacturer specifications as to the -3db low point as absolute gospel.
These specs have not been independently verified with modern measurement equipment. In all likelihood these kinds of claims were massively exaggerated.