xslt critique

http://www.cs.uu.nl/~visser/ftp/LVV02.pdf
has an article about strategic programming, included in it is a critique
about XSLT, I was interested in what people on the list might feel about
these criticisms?
I'm finding some of the complaints not sensible although this is
probably because I have trouble at times with too much abstraction.

Quote from article:
"At least two ingredients of strategic programming are truly missing.
There
is no notion of partiality. Templates simply apply or not. Any kind of
backtracking model is not available. More seriously, XSLT templates
are not first-class. As discussed for basic rewriting and tangled
traversal
in functional programming, this leads to tangling traversal control
and computation. In fact, XSLT even lacks appropriate type-specific
operations since one cannot even easily pattern match on elements of
a certain structure. This was observed in [28]. The resulting
inappropriateness of XSLT for intentionally generic document
transformation is discussed and illustrated in [41]. Besides, an
XSLT-like language design
is hardly accessible for typeful (generic) transformations, although
there
are some theoretical results on typing essential fragments of the XSLT
expressiveness..."


 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list