In Canada, one person died and several people were injured when the stage went down at Bluesfest in Ottawa last July, and one person died and more than a dozen were injured in 2009 when a powerful windstorm caused the main stage to collapse at the Big Valley Jamboree near CamroseAnd this is not like commercial fishing or patrolling in Afghanistan. In those scenarios, death is not only possible and likely, but it's statistically inevitable. Does it have to be so for temporary concert stages?
Six people died last August when the stage collapsed at a Sugarland concert in Indianapolis; five died in Belgium when a storm swept in and toppled the stage at the Pukkelpop Festival.
I think part of this is an engineering problem. We may think these things are designed to serve as stages, but it's only going to live a very small piece of its life as a stage. The rest of the time it will be disassembled components, stacks of bars, bins of bolts. From an engineering perspective, the stage aspect is relatively easy to accomplish. You need an elevated flat area and something from which to hang the rigging. That's it.
But to accomplish that and to break it down, load it on a truck, and set it up somewhere else, that's where the engineers start to wear out their pencils. Add trap doors, pyro, an understage area, the whole thing gets more complicated.
No matter how clever an engineer you are, though, you're still going to be limited by the requirements. It's still going to be constructed hastily, not really anchored to the ground, not rated for freak windstorms, etc.
I think at this point, evolution is going to take over. Either some company is going to design a stage that won't collapse, even in the worst possible situation, or no one's going to want to mount one of these shows. I'm pretty sure Radiohead are not too happy their stage collapsed. Relieved, perhaps, that it didn't occur during their performance, but that's small consolation.