I would say that the software problems are relatively simple compared to the hardware issues of space flight. Software isn't going to offer much improvement to the ballistic facts, nor is it going to make much difference in overall reliability of the in-flight systems - fault tolerance and fail safety have been understood for a long time. I suspect that most of the impact of software would be on the manufacturing side of the project.
To put it another way, once the rocket is lit, the computer is pretty much going where the rocket is pointed regardless of what software is running.
On the other hand, the idea of open source space hardware is pretty interesting and might have a significant chance of impacting the fortunes of people living in the developing world.