The results are finally in:
http://www.iihs.org/ratings/summary.aspx?class=58
The CX-5 didn't do so well, only a marginal score. [...]
3. I wonder how much of an impact the few extra pounds of the AWD system in the test car had on the result. (maybe enough to cost it the acceptable rating?)
I would contend that a CX-5 without the extra 150# mass of the AWD, and the slight reaction lag AWD has, would completely MISS a number of these accidents!
It's unknowable to put a percentage on this crash avoidance, but a FWD is much more lively than an AWD. Today I carried a couple new boxed Ikea desks home, 75# each, in the back. That's almost exactly what the weight distribution would be for AWD, though an AWD's center of gravity would be an inch or two lower. (If you're not shopping at Ikea, stick a mother-in-law in the back seat.
Movement axes are roll, pitch, and yaw. A FWD CX-5 can yaw faster whether with or without electronic stabilization kicking in, simply by virtue of its reduced mass. (Electronic stabilization would kick in after, and in reaction to, the first motion.)
Another way to think of it, at least in the CONUS, is in terms of squirrel avoidance: a small animal runs across the road in front of you. With my old GMC Yukon's suspension, any driving reaction would be likely fatal; hence (straight-ahead) roadkill. My 370Z (aka Nissan Fair Lady) would have more than enough time to avoid, and the precision to thread a couple dimes on the centerline immediately after. The CX-5 is closer to the Z than the land barge despite being 8.5 inches off the ground to begin with.
So: let's say a FWD CX-5 is going to miss 10-20% of the crash scenarios the IIHS has prepositioned for this survey. Should that improve its overall score by a similar amount?