Right - I think I will only try mod-1 (panel sealing) for now.
It's easy to test, and has fewer variables. Spats/flow deflectors will have a ton of geometric variables to work out - so I'll keep that for a future project when I can do lots of iterative tests on different shapes (I might even go as far as to look at a scale model and a mini wind tunnel, though, that is a lot of effort...).
For now though, I'll try taping the panel seams and see what happens. Going to try to get access to an old derelict industrial road with about a mile of straight ways and zero traffic - failing that, It'll work something out