The other issue is that the Cirrus is an unknown. There just isn't enough
of them out there to get good statical data. Of course, the insurance
industry will error on the negative side for you (positive for them.)

How many planes, and flight hours, do you propose is necessary before it is
statistically relevant?