The problem with doing stress calcs like this are many. Sounds like what Hank is talking about is doing simple Mc/I calcs on wing spars and comparing that to published ultimate strength values. Well, that works OK if you have a simple spar in bending made out of the exact stuff that was tested to come up with the properties - but that is rarely the case. You REALLY need to build extra spars and do some destructive testing to see what you really have. And to be statistically meaningful, you are gonna have to test a lot of them. We pretty much never have a simple idealized spar in bending. We normally have some compression of the wing root that helps take some of the bending loads too. Then there is the issue of a spar section's stability failure (on the compressive side), that could drastically weaken a spar. If all you are doing is looking at a published Ftu, then you have totally ignored that possible failure mode (for thin spars only). Then there's the issue of estimating the loads on the spar: not always straight forward! You are assuming a max G load, assuming an aero center locale, and so on. Heck, we have an entire Loads & Dynamics group just for this and NASA's Level II double checks it all independently!

I am not saying that you can not do an accurate stress analysis of the spars, all I am saying is that it is easy to do it, but not easy to do it accurately. You have to know a lot of inputs to a high pedigree for the results to be accurate. Otherwise your errors are compounding errors. Usually, however, the errors are conservative and the answer will show that a spar that works just fine in real life will fail prematurely in flight.