There could be need for nested scan rules?
This could be achieved by a new notation, ++link, which would mean this is a
conditional rule for the previous level, ie the last + (or -) statement. Then
it would be possible to program scan rules like

+download_links_for_subject1_files*
++download_links_for_*.zip*
--download_links_for_*.exe*

This would permit .zip files from only the pages which belong to "subject1",
yes? 
+download_links_for_subject2_files*
--download_links_for_*.zip*
++download_links_for_*.exe*

And .exe files from only "subject2" group.

This way one could do in one scan what would otherwise require two scans, if
the download links themselves are identical, like download.asp?id=78, so no
differentiation on that level is possible. (Getting complicated though..)