Unconventional does not equal non-functional. Burt Rutan designed many of his famous aircrafts precisely to challenge the conventional thinking in aviation design. I remember him saying that most of what we call "engineering design" is actually " engineering analysis" - not design.
Nope. What is at work here is precisely what the NACA duct is designed to do: varying the fluid pressure and flow velocity by varying the chamber volume. As the equation (assuming isothermal) describes:

P1V1=P2V2

when volume is increased, pressure is correspondingly reduced. At the same time, flow velocity also decreases. The decrease in pressure is what holds the vehicle tighter to the road. The differential pressure between the top of the car and the bottom of the car literally sucks it to the ground - increasing downforce on the tires. Those big lateral extensions on F1 cars house large scale versions of these "suckers". Of course, there is no free lunch. Such devices add drag to the vehicle, as it takes forward energy to produce the tremendous down force. Such tradeoffs are advantageous in a racing vehicle. Some production vehicles also employ such "ground effect" devices, notably the likes of street legal Ferraris. A few others actually use ground effects to correct body shapes, which otherwise would produce unacceptable lift at high speed. The early production Audi TT is an infamous example. Ferdinand Piech, VWAG boss at the time, was adamant about not varying from the flowing line drawn by TT's designer, Freeman Thomas. The result was that Audi engineers ended up having to use ground effects to produce sufficient negative lift at autobahn speeds.

Some literature describes vortices generated by the NACA duct sidewalls as contributing to the suction effect, when in fact, it is far from being the principal action. The vortices actually adds to the overall drag produced by the duct.

Well, why not? The equation works both ways. The NACA duct in reverse can produce roughly the opposite effects. Decreasing volume will increase static pressure and increase the fluid flow velocity. It is this increase flow velocity that Mark Langford counted on to help cool the exhaust pipes. Apparently you also did not understand why it is that Mark says it would also reduce drag. Here's the scoop... literally. The cooling air enters the fixed-size intakes on the front of the cowl at the cruising airspeed. The cooling air expands somewhat inside the cowl and slows down. If this air is purged out the cowl through a substantially larger opening at the bottom of the cowl, the slower exhaust air mass will enter the faster moving air stream and cause turbulence (i.e. drag). One way to minimize this drag is to reintroduce the cooling air (now hot exhaust air) back into the airstream at closer to the same velocity. Such is the reason for employing the NACA duct in reverse, which conditions the exhaust airflow by increasing it velocity steadily before re-entering the slipstream. Is there a price to pay? But of course! There always is! The increased static pressure build up before the NACA duct can potentially impede airflow out of the cowl. No doubt Mark calibrated the exhaust ducts to the inlet opening to ensure more than sufficient cooling air flow. He even said he added the center duct for good safety margin and can block it off if later dtermined to be not required.

It is a sound use of the NACA duct - one that shows good understanding of the underlying principals at work.