After a short hiatus (because surfing on my iphone sucks) I've racked up 40 hours of CFD time in less than three days.
As always, the first 27 N came off pretty easily (it only took a single test) and the last 6 N came painstakingly (11 tests).
I raked the spoiler at 45°:
2.5" vertical height
Cd 0.303
Fz 205 N
1.5" vertical height
Cd 0.303
Fz 242 N
Again, 2" height seems to be the best compromise - and I don't expect a "magic" number that significantly improves on that.
My last test, I radiused the edges to 3".
Cd 0.300
Fz 188 N
I reduced the Cd by a full 27 counts over baseline
This is an 8% reduction, which should reduce fuel consumption by 4%. It also reduced lift by 56 lb - 57%. I can't predict the drag and lift improvements for the real car.
I don't know if it's significant, but the spoiler width is almost 61.8% the car's width (golden ratio).
Until I buy a faster, multicore computer, which will allow me to improve the resolution of my results, this is where I'm stopping, as far as the spoiler.