The third problem he raises actually goes like this.
Space flight is too complex and costly to succeed as a private enterprise project at this point. Corporations in competition do not share research and engineering resources, leaving each one to reinvent the wheel (or use old technology).
I will admit that my first reaction involved a snicker. Not very nice, I know, but I know from experience that many of the IT companies born out of the computer industry do indeed cooperate. By day I am a software engineer and I know that many companies will cooperate in order to create a community standard as it reduces their costs. The success of some of the open source and free software code bases is one demonstration of this effort. Look up the project history for Apache for an excellent example.
Would the same thing happen for space flight companies, though? Are there standards from which we would benefit if we took the trouble to cooperate long enough to set them up? From this perspective I should not snicker as these are open questions. I'm pretty sure there are issues where we should cooperate regarding regulatory advice to the AST (yes... they do actually ask for it), sharing of information regarding political and legal hurdles we all face, and in giving the appearance that we can distinguish between credible business entities and the wanna-bees who make up a good web site and stir up some temporary excitement.
It remains to be seen whether we all reinvent the wheel or cooperate with each other to avoid the repetition. If there are enough of us, I imagine we will cover the range of possibilities. Lack of resources will knock some of us out, but a distinct product or service might save someone too. I have my fingers crossed, but it would seem that my faith is that we will win through in the end.
Lastly, my thought regarding the complexity and costly nature of space flight is that it can't be all that hard. We aren't faced with the same political risks we faced decades ago, so there IS room for failure and experimentation. There is no doubt that orbital speeds require a lot of energy, but I know the numbers. Energy isn't the issue. The difficulty comes from trying to deliver all of it in a few short minutes to reduce gravity losses. It is the high power numbers that are costly and complex. We can work very hard on the engineering limitations, or we can work hard at lofting feather-weight systems, or we can suck it up and accept the gravity losses. I know what I prefer and our numbers suggest the technique should work, but all those options are available to someone willing to explore them.