Benefits:
- More approachable to mess around with for newbies -- download the example and try to break it.
- More likely that you'll get submissions from experts who have time to implement the underhanded part but may not get around to putting together the boilerplate.
- More likely that submissions will get the general task right and look similar for parts that don't matter, making for easier review.
Drawbacks:
- Potential for too many cookie cutter submissions.
- Might lead people away from thinking about alternate server designs that are more underhanded.
Mostly I think this might be worth doing because of the second benefit -- you're asking people to do a bunch of grunt work even after they've figured out the core underhanded technique.