> download, it re-downloads every PDF, every image,
> etc and it takes hours to do so.  Seems like it
> ought to be just downloading changed or new files. 

urls in the form of page?args=... are server side code and the html output are
new each time and therefor downloaded each time.

For the PDFs have you tried options -> Spider -> Update hack
Also make sure HTTP/1.0 is not checked