You could try parabolic dish antennas. Boith JPL & MIT have studies on this which I suspect are public domain. However they don't necessarily agree for all cases. I ran a bunch of these consditions for a client and we were alwasy within 9% of the one set of data or the other. I can't release anything but if you can get the data/papers it should be a pretty easy excersize to pull off.
I've had great success wth FloWorks and am usually within 10% of actual.
With futher mesh refinement, you could probably get within 5%.
One of the problems with EXTERNAL structures and WIND is that FloWorks cannot account for flutter and non-linear wind loads.
In your case, you're probably going to need to import the pressure/temperature (as static) loads to do a structural analysis. Which should yield good results.
All of the SolidWorks Analysis products have verification problems on which they test the solvers agains real life known problems. Those should be sufficient to prove that the tools accuracy.
That being said, ANY analysis is NEVER going to be 100% accurate and doesn't replace real world testing. It is impossible to account for everything (defect in material, welds, etc). BUT, with analysis software the answers you get are usually in the right ballpark and is USUALLY conservative.