So, I thought I'd pen a slightly longer response.
First, I should note that I believe you and I agree on the ends (advancing technology) and may only differ somewhat on the means (e.g. primarily driven by USG vs. a more global strategy). What DOE in particular and USG in general is doing with things like GTL or fusion research is fantastic. But USG has a lot on its plate and can't do everything.
Second, I believe this paraphrases your argument compactly (and please tell me if not):
VCs don't invest in fundamental scientific innovation, and for big breakthroughs we need public/
private partnerships. It is unrealistic to expect software engineers to make breakthroughs in
physical energy and transport systems; these kinds of investments require billions of dollars and
timeframes beyond VC funding horizons.
In response I would make a few points:
1) Would fundamental innovation in science happen without USG support? Obviously, USG-funded research led in part or whole to the Human Genome Project, space travel, the atomic bomb, the computer, the Internet, and (most recently) the self-driving car. So USG support is certainly a sufficient condition.
2) However, it is not a *necessary* condition. After all, Newton, Maxwell, and most of science and engineering predate the establishment of the National Research Foundation and Vannevar Bush's 1945 memo.
3) In this sense, all of pre-1945 fundamental innovation happened without USG support. This happened in mathematics, like the development of calculus (Newton & Leibniz), algebra (Al-Khwarizmi), and the zero (Aryabhata). But it also happened in physics: the applied science of steam engines led to the pure science of thermodynamics and later statistical mechanics. See:
4) Today the applied science of software companies is leading to a flowering of pure science in the form of open source. One might argue that the latter is simply "tool development", but much pure science is about tool development (viz. the Nobel Prizes for PCR or the Scanning Tunnelling Microscope). And open-source technologies like Linux, Firefox, Bittorrent, nginx, and the like are at the very least *extremely* sophisticated tools.
5) In the past, most scientists were independent "gentleman scientists", similar to the open source community today, rather than a professional class of grant-funded researchers associated with the academy. Now, I would be the first to say that the grant-funded model has achieved many great things; however, it is not the *only* model that has stimulated innovations in fundamental science. Again, there are large numbers of people on lists like this (http://en.wikipedia.org/wiki/List_of_physicists) who were born before the 20th century and/or not funded by USG. And virtually all of the open source community's work is not funded by USG.
6) The broader question of how to accelerate fundamental innovation is a deeply important topic. Jerzy Gangi has a good post on "Why Silicon Valley Funds Instagrams, Not Hyperloops" which I believe you may find enjoyable.
7) I agree with you that VC as currently constituted is not set up to fund fundamental innovation without near-term monetary return; this does not mean it is impossible to envision third structures different from both USG and VC (like open-source).
8) All this is related to, but separate from, the subject of my talk. You mention Elysium and the Matrix. I would offer a different analogy: the Inverse Amish.
Just like the Amish live nearby, peacefully, in the past - imagine a society of Inverse Amish that lives nearby, peacefully, in the future. A place where Google Glass wearers are normal, where self-driving cars and delivery drones aren't restricted by law, and where we can experiment with new technologies *without* causing undue disruption to others. Think of this like a Special Innovation Zone similar to the Special Economic Zones that Deng Xiaoping used to allow China to experiment with capitalism in a controlled way.
9) In sum: I believe that regulations exist for a reason. And I believe that new technologies will keep coming up against existing rulesets. I don't believe the solution is either to change the rulesets (which, again, exist for a reason) OR to give up on new technology. I think instead we need a third solution: a way to exit (whether to the cloud for purely digital technologies, or to a Special Innovation Zone or ultimately a startup nation), prove/disprove these new technologies among a self-selected, opt-in group of risk-tolerant early adopters, and report back to the mothership on what works and what doesn't.
10) This concept - a Special Innovation Zone - is a new idea. It is really about humility, not hostility. USG is a big thing, it has a lot of responsibilities, it runs a nation of 300M people, and it can't just change federal laws to permit some crazy tech guys to try (say) self-driving cars without affecting millions of people. A new region - like a Special Innovation Zone - can experiment with this kind of thing without bothering anyone who wishes to live under the previous rulesets.
Again: this is complementary to USG's own efforts. I don't see them as competitive, anymore than a startup competes with IBM's research labs.