It was getting the "Load More results" to work correctly that made it take as long as it did! It is now working fine for me, so I would like to know why it is not working for you?
I uncovered a bug (reported to Lexcycle) where Stanza 1.7 will crash if for any reason it fails to get the next page of results. I originally uncovered the Stanza bug by making the link to the next page point to the wrong file. Perhaps you can check that each of the "Recent#.xml" type files generated by the script on your machine has a "next" link that correctly points to the next one - with the case of filenames possibly being significant if not running on a Windows web server.
As a work-around you can change the PAGE-SIZE constant in the script to be at least as large as the RECENT_ADDITIONS constant so that all results fit on one page.