Of course an idling prop doesn't windmill. But at some rpm an engine on idle will be equivalent to a windmilling prop. And of course there is always some friction involved, so obviously the rubber fliers prefer a prop turning at some speed that creates drag, but less than what is created by a prop in a stalled condition. I didn't really expect to have to explain that.

So picture an engine running at low rpm, not producing thrust, equivalent to what it would turn if it were windmilling in a low friction condition. Now slow it down just a little. Obviously it will produce a little more drag than at windmill speed. Is that amount necessarily more than the drag of a stationary prop in a stalled condition? Of course not. One can slow it down slightly and there will be more drag than when it windmills, but less than when it is fully stopped.

Now slow it down some more. More drag? Well, if it is still not in a stalled condition, there will be more drag. To use your words, more "negative lift". Now, it may be that a stopped prop in a stalled condition produces less drag than a moving prop in an unstalled condition if the rpm is low enough. But not at ANY rpm below what is required to produce thrust.

Also, at some rpm, a prop turned by an idling engine will nevertheless be in a stalled condition. Is a prop that is turning in a stalled condition draggier than a prop that is stationary in a stalled condition? I don't know, but I can't see why it would be.

I am willing to believe that under the right conditions an prop turned by an idling engine could be draggier than a stopped prop, but if so, it will be true only under particular conditions that depend on the prop characteristics, including its pitch, and the speed of the plane, and will only be true through a particular rpm range that may be much narrow than the range we consider to be at idle.

Jim