Assuming a 100 foot tall building, that would be a building 400 miles on each side. For comparison the amount of solar energy delivered to that same area is 20 times higher! (Assuming you could capture all of it.)
http://www.wolframalpha.com/input/?i=%283.3991%C3%9710^12+wa...
So to make fusion practical for human use it's not enough to duplicate the sun - they actually need to make it work far better than what happens in the sun. That's one of the reasons they use D-T reactions instead of H-H - it's much faster.
I always believed they were trying to duplicate the suns core, not the whole ball of gas.
Much smaller - still pretty large, but doable. Your energy output in comparison to capturing solar energy is now 3 times as much rather than 1/20 as much.
Article in Physical Review Letters: http://prl.aps.org/abstract/PRL/v108/i16/e165004 (paywalled)
Freely accessible preprint: http://www.pppl.gov/pub_report/2012/PPPL-4728.pdf
I mean, there is always a problem with oversimplifying these things, but talking to a layperson doesn't have to be too much different from talking to a fellow physicist; the key difference for me seems to be presenting two or three perspectives on the same thing before going forward.
Here is the backstory to this paper, as I would explain it to laypeople: "So we've got this plasma -- plasmas are like a metallic gas, lightning is a good example; the electrons get shared between the gas molecules so the gas becomes an amazing conductor. Same thing happens whenever we superheat any gas, really. So we've got this superheated gas in a tokamak, which is a big ring which uses magnets to keep the plasma spinning around in a circle without hitting the edges. It's a neat trick, a bit hard to do. Now if we could get the plasma hot enough and dense enough, we could recreate the conditions in the Sun with it, and solve the energy crisis -- this is what's called fusion energy, much cleaner than nuclear fission. But how do we heat it up? Well like I said, this plasma is metallic, it's just like the wire in a light bulb, so we just put a lot of electricity through it, and it heats up and glows. We call that "ohmic heating". But there's a problem. Before it gets hot enough for controlled fusion, we basically run into a wall, we try to pour more current into it and it doesn't get hotter, doesn't get more dense.* And until now we didn't really know why this happened. It's as if the laws of nature just said, 'Eff you, I'm not letting you solve the world's energy problems that easily.' We have an explanation, now, why..."
I haven't read the paper yet but I hope it can rise a bit to a better technical level than the press releases. The press releases are a little elliptical about what their solution entails but it sounds like you have to understand what's happening with certain observed impurities in the plasma.
* From what one of the press releases says it sounds like it might actually kill the containment field and fly apart, but that sounds really strange -- as if it would destroy the tokamak.
It's _really_ hard to keep the magnetic field right. It's not a stable system, it needs to be dynamically controlled. Presumably these defects mess up the magnetic field to the point that it's impossible to contain it.
>"Among other things, they intend to see if injecting power directly into the islands will lead to higher density. If so, that could help future tokamaks reach the extreme density and 100-million-degree temperatures that fusion requires."
