PostFrameSE:
I don’t know that that is something which is really calculable. But, do a good job of laying out and thinking through the chain of events which allows such a thing to happen and you probably come to the conclusion that is doesn’t take hurricane force winds to cause these roofs to take flight. If they comply, they barely comply, with the IBC, IRC, or the NDS, you know the saying, ‘save a spike or a bolt, and yu save a few minutes and a buck or two.’ They are relatively inexpensive farm buildings with little likelihood of much public safety involvement, so the codes and AHJ kinda let them slide. At first you have some wind pressure on one side/surface and suction on the other side, partially open building probably, over a large contributing area as relates to truss connections on beams or at posts. Don’t forget that pressure differentials and fluctuations are highest at eaves and building corners where the ripping apart invariably starts. Once you’ve caused/allowed the hold down connections to fail, and uplift which was probably never designed for in the basic design and detailing anyway; then it only takes a positive/uplift pressure one percent over the weight of the roof system for it to take flight. Then what happens, who knows, a slight gust of wind, in a slightly different direction, and it goes 400' instead of only 200'. That would be my explanation of what happens. If you want to go a little deeper take a look at the design load, wind load, wind pressure criteria in ASCE 7, and take the weight of the roof structure above the truss bearings and you should be able to back out a wind pressure or wind speed from their convoluted approach to wind loads to make your roof fly. Their approach has considerable testing behind it to make us much smarter about how this all happens, but it is still not an exact science. The big question will still be what forces allowed the connections to fail in the first place, and get this all started.
