Still thinking about longitudinal oriented static magnets in the pavement-- what if you synced the car's engine and recharging system so the engine is only working during the first half of the car's traverse over it, but then turned off the motors and coasted over the second half? Then you'd be recharging without expending any energy from the motors at a 50% duty cycle.
If the magnets were spaced at short enough intervals it's possible the acceleration/coasting wouldn't be very detectable. Obviously, if it WERE detectable, then it wouldn't be much fun. It would feel like how my FIL drives, which almost makes me sick to my stomach sometimes... 
The more that I think about it, I don't think longitudinal can work.
I feel like you're looking at the alternating magnets like a pull/push situation, where the actual behavior is a "resist the change" function.
Think of the passing a bar magnet through an electric coil example. To think of this similar to a car, assume that we are working in the frame of forward/backward. I.e. you've got the bar magnet on a string and you're going to pull it TOWARDS you through the coil which is mounted in line with you. So it will simulate a vehicle moving forward through the magnetic field.
When you try to pull the north end of the bar magnet through the coil towards you, the electrical current induced will create an opposite magnetic force to the movement. I.e. it will create a field pushing against the magnet, but because you're pulling the magnet through and it can't stop it, the force will be applied to the coil in the backwards direction, i.e. towards you. So in the car scenario, this would be a force against the direction of travel.
Now, you've fully pulled through the magnet so the north end has exited the coil and the south end is moving through the coil. So again you're changing the magnetic field. In this case the current in the coil will be reversed, and it will create the opposite magnetic field attempting to stop you from pulling the magnet out. Because you're pulling the magnet and it can't stop you, the force will be applied to the coil in the backwards direction, i.e. towards you, AGAIN. Thus in the car scenario, this would AGAIN be a force against the direction of travel.
It doesn't matter whether you start with north or start with south, the forces will always be opposite motion of travel. The only reason to either make the magnetization intermittent, or to make it alternate, is to make sure that you're constantly changing the field because you will only generate current when the field is changing.
But in every case, the force is applied opposite the motion of travel when done longitudinally.
That's how I remember it from all those physics classes, anyway ;-)
