Don Box stopped working on an XSL transform of WordML to XHTML.
Microsoft WordML to HTML XSL Transformation appears to have vanished from the Download Center and is probably buried in the MSIE-only Word 2003: XML Viewer (wmlview.exe). However, this guy, Oleg Tkachenko, has his own version of the XSLT file here:
http://www.tkachenko.com/dotnet/files/Word2HTML-1.0.zip
He journals the mysterious behavior of Don Box in "On transforming WordML to HTML again" here:
http://www.tkachenko.com/blog/archives/000153.html
What's left out there are expensive tools that convert entire files to XML and even a web site where DOC files can be uploaded and converted to DocBook format. Most of these designs avoid intimate contact with the automation features of Word. Most are one-way tickets out of the Word format. I need a bit more flexibility.
John E. Simpson, surveys the tools out there in "From Word to XML" here:
http://www.xml.com/pub/a/2003/12/31/qa.html