Not to disagree with you general point, but applying general risks to specific situations is not proper use of statistics. Once a thunderstorm is on, the risk just multiplied from the baseline risk for the people in the area. And it grows again for people who take a shower.