I think we pretty much agree on all this, particularly the effectiveness of the hybrid model. >What we are seeing is actually the commercial company are far mo

I think we pretty much agree on all this, particularly the effectiveness of the hybrid model. >What we are seeing is actually the commercial company are far more willing to risk these things the governments. My personal opinion is that that NASA (less so the military) are risk-adverse in terms of political risk. When there's little political risk, they are quite fine blowing stuff up to learn. One thing that I've noticed is that commercial companies tend to take an engineering perspective and NASA tends to take a more fundamental science perspective. For example, when SpaceX had COPV issues, they were fine to just change the design and move on without understanding the root failure mechanism. NASA funded their own studies to try and understand what was going on. This is the direction I'd like to see the hybrid model continue. I do think that contract changes can risk efficiency drops, though. Right now, CCP doesn't have to show a lot of their work. This may mean that their "efficiency" is a product of process gaps, rather than better processes. The CST-100 software validation was a prime example. It was a definite process gap that allowed them to get to launch quicker. If they were lucky and didn't run into issues, NASA would be none the wiser.