I'm not sure what you would be trying to discover by measuring the heart rates of the players, but that does sound interesting.
I think the biggest problem with doing an experiment like this is controlling for the (presumed) fact that boys typically play many more video games and for much longer than girls, and therefore might just be more practiced at them in general. Even if you have them play a game that none of them have seen before, practice at one game probably carries over to others.
A really interesting experiment, and a way to control for this variable, would be to find both boys and girls who have never (or only rarely) played *any* video games (good luck
and see how much they improve with a certain number of hours of playing. I've heard of a study like that and it had really interesting results.