Everything you said is true, but the addition of this adds obscurity to the XPath selectors. Namespace confusion doesn't really exist as the entire document is written in the specified namespace.
I suppose another dirty way is to strip the xmlns attribute from the document before parsing it... :-)
Anyway my basic question has been answered in the negative, so I'm going to simply have to live with the "clean" way of doing it....
A Monk aims to give answers to those who have none, and to learn from those who know more.