I think you have to consider the opportunity cost. Spending $100k to kill a $500 drone sounds like a bad deal, but if that drone kills a soldier it could cost hundreds of thousands (or even millions) to train a new one and pay out death benefits.