Originally Posted by
msbbarratt
Getting every component exactly right every time is a ferociously expensive business.
Except they don't. As evidence we have two shuttles lost in only ~130 launches, and infamous process errors like the tool left in the engine bay even though three people had signed paperwork to verify that it had been removed.
I'm far from convinced that efforts like Space X can make truly large cost savings in the satellite launching business.
In the long run, the only way to make flight to orbit cheap is to make the launchers reusable with minimal servicing. The question is whether you start with something that isn't reusable (SpaceX), and work up, or start by trying to build something that is reusable from the first launch (Skylon).
Skylon's problem is that, like all current single-stage-to-orbit-designs, the payload is small, it could easily become negative as the vehicle mass grows during development, and you have to invest billions of pounds up front before you can find out whether it's possible.