I would also be more than happy to donate all 200mbit of my available bandwidth to this effort. Maybe some folks who are much more talented at scripting than I can figure out a way to scrape a list of URLs to the PDFs.