The empirical evidence certainly doesn't suggest shorter stages mean less drugs. As Lim pointed out, in the early 1900s, stages might be 400km long (and start at midnight, and be on terrible roads, and with bikes that had only flip-flop rear gearing, and no support cars, and they had to carry all their gear, and they couldn't even drop anything!) and have only gotten shorter. Less doping? No way.
Sprint runners like Carl Lewis and Michael Johnson dope. Dope doesn't care how far you have to go. It's still dope, just maybe a different kind.