if the Statistic function is broken or not is really a question for the bug report forum. Earlier with your sphere I showed that the value it reported was correct, because of how triangles relates to polys (2 to 1) but this may not always be the exact number, it really depends on the mesh in question and how it was modelled. 2 to 1 should be used as a baseline only.
If I make a sphere in XSI, 36x36 UV, and triangulate it I get around 2500+ triangles, which is about right (36*36*2).
I would just try to use as little polys as possible within acceptable limits to the eye, and make your game, and see where you start to notice where the polys begin to overtake the engine. "Triangles" are the concern of the rendering pipeline, and for bloated advertising purposes, for video cards
If you are concerned about the function then maybe search/post in bug forum. Theres probably more to it than I have been able to explain - but I have relayed what I know.
EDIT: Your asteroid.x model reports 11004 triangles in SI:XSI 6
EDIT: Your ast4.x model reports 5880 triangles in SI:XSI 6
My DBP plugins page is now hosted [href]
here[/href]
