I'm not sure if I agree with the conclusion of the article; if the problem is that the effect of drag increases with speed, then the answer here seems to be to decrease drag (perhaps with a more efficient flight profile). Or would that interfere with lift?