I wanted to get a better view of the effect of the box lid-top radius, this time I added a piece of cardboard over the foam to recreate a sharp edge with the same tangents as the radius version, that way I don't have any change in gap size as a confounding factor.
I think this best fits the null hypothesis - radius at reattachment does not attenuate the locked vortex. (I can't wait for kdenlive to get image stabilization, youtube's stabilization doesn't recognize stills. EDIT; actually kdenlive does have stabilization - just a little counter-intuitive.)
It's nice to get this detail sorted out before fab. What this means to me overall is that the version most worth building should close the gap as much as possible. Also there's no point aerodynamically in rebuilding the box lid. Also for icy weather I may need to have heater lines in the rear window (since the alternative, a removable window, would open up a large gap and potentially negate the aerodynamics of the Kamm section).
There's also a better view of the underside at the end of the vid. To me it looks like it's begging for a belly-pan and wheel treatment.