Well, I dusted off the textbooks and the right term for what I have been trying to express is "polar moment of inertia". In my case I confused the term with RG. The polar moment is a different mathematical expression from what is commonly called "moment of inertia". So, we are all guilty of using less than accurate terminology.
The assertion that Tornado hulls have very little weight variation came from a 1996 article in Cat Sailor magazine that was describing preparation for the Olympics in Savannah. The measurer stated he was surprised that weights of the Marstrom boats only varied a few ounces. But, if the measurer's claim is still suspect, there is data from many other composites applications. For instance, carbon helicopter blades completely out perform aluminum and have better tolerances on both weight and stiffness.
It would not be practical to alter a carbon stick by sanding it. You could not remove material with enough uniformity to achieve any desired end result. Whatever you ended up with would be sheer luck. If someone really wanted to cheat, it would be much more feasible to chemically mill an aluminum tube than sand a carbon one.
There is a potential problem with measuring stiffness as a way to ensure consistency. This is because even if stiffness of both the major and minor dimensions were measured, games can be played by altering ply lay up alter stiffnesses in other orientations. What makes the unidirectional fiber such a flexible material to design with becomes a liability when trying to assure exact properties from one builder to another. To remove variation in composite sticks it drives you towards a single builder. I reserve judgement whether this is good or bad. It's certainly not in the spirit of the original Tornado rules but then the current aims of the class are different.