To sum it up "Free energy" buys us about 300-400 years of growth before thermodynamics kicks our ass.
That growth will stop sometime and a considerable amount of present pain seems to be that global growth is slowing markedly (and has been since the 1970s).
Globally, economic activity and energy consumption are still tightly correlated: http://www.reddit.com/r/dredmorbius/comments/1vlksg/economic...
It would be "strange" to argue that, as a teenager, you were doomed to a life of misery because in order to keep up with this so-far exponential growth rate you'd have to eat entire planets by the time you were 1000 years old. It is similarly bizarre to argue that human society is in some sort of trouble because we can only double our energy consumption a few dozen more times.
This isn't even bringing up economic concepts like diminishing returns to utility: a person who makes 100k per year does not spend twice as much energy as someone who makes 50k per year. There are plenty of other things to buy that make us happy, and that can only get even more true when we're living in some hypothetical future with hundreds of times the economic productivity of today.
On the other side, Robert Gordon's "End of Growth" paper: http://www.voxeu.org/article/us-economic-growth-over
Your assertion that a person earning 100k doesn't have twice the energy (or environmental) footprint of someone earning 50k really needs some sort of factual support.
And your hypothetical future with 100s of times today's economic productivity flies in the face of the law of diminishing returns which you've invoked yourself.
http://en.wikipedia.org/wiki/Economics_of_nuclear_power_plan...
Fusion reactions can only be self-sustaining at incredibly high temperatures and pressures. If the containment mechanism loses power, the reaction stops. Unlike fission reactions, at least as used in current reactors, which have the potential to keep on going unless we do something active to stop them. Yeah, the modern, well-designed ones have lots of designed-in safeguards against this, void coefficients and all, but it's still safer still to be running fusion.
Hydrogen and it's isotopes in, with other isotopes or helium out is better than assorted radioactive heavy metal isotopes with relatively long half-lives. The only possibility for dangerous waste from a fusion reactor is any transmutation that takes place in reactor components from the neutron flux. Seems likely to be better, but I don't think anybody has really investigated what this part will look like in a running fusion reactor with reaction rates high enough for commercial power production.
http://en.wikipedia.org/wiki/Fusion_power#Waste_management
A lot safer too!
http://en.wikipedia.org/wiki/Cost_of_electricity_by_source
What I don't know is how much of the cost of nuclear power is the much higher level of safety precautions they need.
I haven't seen any projected costs for DEMO which might well be the first fusion power plant (DEMO is the proposed successor to ITER):
It has to be, otherwise nobody would be using nuclear power plants...
Assuming we don't use the new energy in significant part to colonize other planets, with the colonies representing a significant part of the growth in human energy consumption.
Water shortages could be addressed with large scale desalination. Shipping/transportation would become real cheap without significant fuel costs. Food issues could be addressed by commercial hydroponics.
I have a feeling that if we had fusion plants providing limitless cheap power, we'd get back on an 8% yearly growth track for energy consumption, like with oil consumption before 1970.
Then within a century or so, we'd be producing an amount of power equivalent to 20% of what the earth receives from the sun. It would give a whole new meaning to "global warming".