This is a really compelling example of how one can be more useful than the other.

Or perhaps just a demonstration that monte carlo simulation is a really bad way to calculate π, and picking random points can make it still worse. ;-)