How loose is too loose?
 

Someone, I think treb but can't remember atm, mentioned that a loose cover for a boat tail could be as bad as none at all, and I was wondering, what constitutes loose? Based on internetz searches, it seems like oscillation may not be a bad thing for separation, especially if actuated, so how much give can a surface have before it becomes relatively useless compared to a rigid surface?

Huh. You mean like a luffing sail, where the vortices are still following a surface? Tricky to model.

I think it would set up one helluva racket inside the car. For my money, loose at all is too loose - there's squeaks and clunks enough without adding more.

Not exactly. I'm just wondering if using, say, thick clear vinyl, which will have some amount of play/give to it even if it was fairly taut, would be significantly worse than something solid.

Not having done anything in clear vinyl, it is a little tough to say. However, my quess is that your not going to be able to get it tight enough to keep if from flapping and when it goes into flapping it will be extremely annoying, at best. I'd recommend using something more rigid, but it just a suggestion.

looseness
 

My thought is that if the shape of the envelope is such that it is "just" maintaining attached flow,that if it moves or deforms to a contour that exceeds what the flow will follow,that flow will separate,defeating the whole premise of the tail.My inflated boattail on the T-100 had an aluminum band that held the envelope to a cantilevered bottom spar,which prevented either vertical or lateral displacement.Air pressure inside the envelope provided enough skin tension that it performed as a "solid" surface.Tuft tests demonstrated attached flow over the entire surface.