Then why shouldn't we apply this same standard to government projects? Any time a government project doesn't go perfectly it's used an excuse to scrap the very concept of government since "they can't do anything right".
As far as I'm concerned, the government is free to innovate, as long as my daily life (and that of my fellow Americans) is not part of what's being experimented with. We have the entire corpus of world history to refer to for experimental data on government policy -- let's use it as much as possible. Social crises aren't worth creating for the data they yield.
Regarding technical innovation, NASA did a lot of that in the 1960s. There were a lot of failures then (Mercury and Gemini programs), followed by incredible success (Apollo program). NASA's work with SpaceX is another great form of innovation -- pick the most innovative commercial partners and move forward.
The US culture around government is really weird. And yes, if people gave the government permission to fail sometimes they could probably do things 10x cheaper. But that requires a level of trust that just isn't there. I don't know why, and I'm glad I live somewhere more functional.
No only does lack of trust prevent innovation, on the other end of the spectrum you have people who won't tolerate failure because it might give the distrustful of government people data point or talking point.
