Marlys,
The problem with both the Doc Morph and translation servers is the
same as with the latest screen reader technologies, and that is that
are still plenty of documents that won't convert into anything
meaningful, if at all.
The translation server you mentioned is at the Trace Center at the
University of Wisconsin, Madison. There is another one hosted by
Adobe themselves.
Actually, with either Kurzweil OCR software or the FineReader OCR
package, we now have slightly more convenient access to some PDFs
which consist of scanned images. Depending on the quality of the
scan, this can actually work quite well. Most of my reading for a
graduate cognitive science course and a graduate readings seminar
consisted of scanned PDFs and with one or two exceptions I was able to
read them fine -- several of them were as good as any OCR I have ever
encountered.
-- Jim