The criticism I've long had of the people who talk about distributed generation is that efficiency is necessarily lower and costs per kW-hr will necessarily be far higher than you would see for centralized generation: for solar, for instance, efficiencies of residential installations are typically half that of centralized facilities.
Yes you may get some advantages from co-generation - but only in locations where you actually need heat at the same time you need the most electric power; under most conditions in the US and warmer climates the needs for heat and electricity are anti-correlated, and co-generation isn't going to make much difference to efficiency.
In Smalley's article for MRS, he outlines the challenge, dismisses the hydrogen "solution" as a distraction, and strongly emphasizes the potential for solar power. He also advocates a continent-wide electric grid, at least across the America's - and then points out that for energy storage and transport we need to be keeping it in some electric, and not chemical, form.
Instead of storing energy in some chemical form, keep it as pure energy. There are essentially only two ways to do that. We could microwave energy up to a satellite and bounce it back down, or we could run it along wires on the earths surface.
But the problem for solar, wind, etc. at the large scale, at least for solar on Earth's surface, is intermittency: we need not just to transport the energy efficiently, but also to store it for overnight use and other periods when the grid may not be available. Smalley advocates local storage, and it may actually work:
Imagine that by mid-century, nanotechnologies, new materials, and possibly new physics will have enabled us to create local storage units for electrical energy that are not much bigger than this lectern. The units would store 100 kilowatt-hours, which is enough to run a normal house for 24 hours.
Smalley may be overoptimistic on this - he claims that even now such a system (with lead acid batteries) would only cost $10,000, though it would be much larger than his "appliance" size suggestion. Having purchased a large UPS system (based on lead acid batteries) for maintaining power to a computer facility, I think that's low by at least a factor of ten in the present market. And batteries wear out - typically over five years at most, or if kept in poor conditions (for example, not air-conditioned) they may give out after only 2-3 years. Replacing those batteries for current systems like this makes such a local storage solution even more costly.
Nevertheless, it's an interesting challenge. If Smalley's proposed 100 kW-hr energy storage appliance could ever prove practical and affordable, he's quite right that it would be a great solution to compensate for inconsistent power from large-scale solar installations and other interruptions to steady grid supply. Why is it so hard to store energy efficiently and affordably anyway?