Funny you should ask this - I've been doing some wind-tunnel tests the last few days to work out a very similar question for drag on the string of a very high balloon. It's not straightforward.

A rough order-of-magnitude answer would be that the aero drag will be similar in magnitude to the weight. Aero drag is roughly .5 * rho * v^2 * A * Cd where rho = 1.2 kg/m3 for air, A = frontal area = 0.4 m2 say for an adult, Cd is the drag coefficient = 0.5, say for a typical bluff body. The man weighs 76kg == 760N so speed v = sqrt(760/(.5 * 1.2 * .4 * .5)) = 80 m/s == 150mph - not far from Soul Surfer's answer based on skydivers, because after all this is how you calculate terminal velocity.

But as Soul Surfer says, it's tricky to estimate what will really happen in the turbulent wake behind a car. Some may suggest that the lift coefficient should be used, but at what angle does drag become lift? And what is the lift coefficient for a human-body-shaped object?

Any use?

Hugh