Thanks for the encouragement everyone...
Yes, there's not much data above 65mph... and it's not clear that the slight increase in flow with the Cup roof is significant above 65... not many data points. I'd love to get data at higher speeds, but as you have guessed, my "course" is on local roads... and I might have problems if I tried to take too much data above 75mph
EDIT: The "fixed course" is close to a square, with approximately equal legs in each compass direction... and I tried to avoid taking data on windy days, but a sudden gust of wind could always be responsible for data points above or below the regression line.
I wanted to make sure I could reproduce the conditions somewhat reliably so that the data could be compared properly.
I've just started "fettling"... I'll post more results when I've got them...
__________________
* 2009 Lotus Challenge Series Rookie Of The Year *
(Thanks Jim, Tom, Jack, Rob & Spring Mountain Motorsports Ranch!)
My videos:
http://www.youtube.com/user/apk919