|The stupid question is the question not asked|
Every coin has two sides.
While most monks rightly criticize the "certification industry", where "simple" people are learning stupid multiple choice tests by heart to impress "simple" HR folks, the learning effect is ignored.
I applied recently for some Perl projects and was horrified to see at what standards those projects where run. I mean those people didn't really know what strict means or how moduls are used. (I ran into a situation where the boss got afraid about my "qualification" endangering his authority. =)
I think having something like the Euler project with a extendable online collection of Perl exercises, maybe cross-linked to online tutorials or chapters in O'Reilly Books (some of Merlyn's book have a training part at the end of the chapters) would be a good thing.
But this side of the coin wouldn't need to include a reliable "measuring" the knowledge in a linear way, to produce a rank for HR-folks.
This measuring could - if ever - only be done for very canonical things which include basic techniques and terminology everyone agrees that should be understand.
(scopes of variables, strict, package, use, type inference (== vs eq),...)
BUT like Monk's XP system is not a reliable measurement of knowledge it works as good motivation.