[Frank Lee]

The difference with a Kamm body is they are able to make a wake zone that has smaller area than the frontal area and the frontal area is determined by the size of the interior they desire. With a mirror housing, you have x size for the mirror and I think making the housing larger in frontal area so that you can taper it back down to the size of the mirror at it's trailing edge would result in greater CdA simply due to the greater frontal area even though the Cd itself might be better. Or if I'm wrong about that maybe nobody does it- not even racers- because they don't like the looks?

I'd think there's an optimum fore-aft length for mirror housings though. Thinking about the mirrors on my '60s Chevy cars and the big mirrors on my '76 Chevy Camper truck that are just about plain flat planes- I'd expect more turbulence off them than mirrors with housings with enough length to have a place for flow to be attached or reasonably linear before it separates off the back, like the ones on my Tempo or Sable.

The clear trailing fairings might have some merit if they can be shaped to not interfere with the mirror's image. Then too there are probably issues with fogging and icing and how do you clear it without scratching it up, etc.

I'd say the best bet is minimizing the size of the mirror if you can do it. Or eliminating the mirror if you can do that. Or locating the mirror in a less accelerated flow region. Maybe pressurized air can be fed to the mirrors for air curtains?