|Perl Monk, Perl Meditation
I have another concept that you might add to your tutorial. Evolutionary algorithms assume that the route to an optimal solution is paved with progressively better solutions. This is not be true for some problems, so it's important to keep that in mind.
Also, sometimes your evolutionary algorithms tend to migrate towards a sub-optimal solution that is not anywhere "near" an optimal solution. In your hill metaphor, you might call this a false peak of sorts. One way around this is to introduce a cataclysmic event that kills off most of your population and replacing them with randomly initialized individuals.
One more thing: at my school we call them genetic algorithms, perhaps you could not that somewhere in your "What Are They?" section?
Awesome write-up, this is a very good explanation of a cool paradigm in problem solving. :)