How about doing a statistical analysis of the nodes here on Perlmonks(and comp.lang.perl.misc), and presenting the most commonly asked questions (master level and not already a faq entry) with the best solutions offered? Maybe a picture emerges.