One of the things in aviation I have always accepted the pat answer for is that a windmilling prop produces more drag than a stopped prop, and that the amount of drag being produced by a windmilling prop in a piston-powered aircraft is roughly equal to that of a flat disk approximately the size of the propeller arc.
Obviously, in a single, when engine failure is encountered, you want the prop to windmill (for air starts and keeping systems going). I've never had an instructor advise me that there's a time to stop the prop from windmilling (it seems like you could do it by bringing the nose up and bleeding off your airspeed until you're near the stalling speed, if need be). The only place I could see this being of value is if you had to stretch a glide, for instance, engine failure over mountanous terrain with no suitable landing site within gliding range.
I won't even question feathering in a twin with a single engine inoperative!
But seriously, any insight into where the drag figures we've all learned in ground school for a windmilling prop come from?