What seems key to me is having a large enough energy output to harness without waste in the flare window while being able to manufacture fuel in sufficient volumes to power the sucker.
Focus fusion would be a small boron fusion device, producing 5MW per reactor at a tenth the price of coal. Potential drawbacks: the energy gain is a bit marginal so you have to also capture x-rays, and erosion of the electrodes could be a problem. But a 5MW plant should only cost half a million or so.
Picosecond laser would also do boron fusion (which in that design is only ten times harder than D-T), but with a 10,000x energy gain, divided by the one percent efficiency of the laser. It's projected to produce power as cheaply as focus fusion.
Tri-Alpha supposedly could achieve boron fusion but they've been very tight-lipped about the whole thing. They've gotten close to $100 million in venture capital though, including from Paul Allen.
Polywell, if it works, could do D-T fusion in a 2-meter sphere, or boron fusion in a 3-meter. I've seen relatively high projected costs, though that may be for D-T. If you're a scifi fan, this was invented by the guy who came up with Bussard ramjets. He was also a big tokamak guy, before he gave up on them.
Levitated dipole sorta turns tokamaks inside-out, with the plasma contained around the outside of a solid levitating torus. This seems to solve some plasma instabilities that tokamaks struggle with. They say it would be ideal for D-D fusion.
Deuterium (for D-D) and boron are readily available in large quantities relative to what we'd need. D-T needs tritium, which would have to be made by bombarding lithium with the neutrons from the fusion reaction. This would limit how fast we could roll out the reactors. D-D and D-T would both require steam turbines. Most people focus on D-T since it has the least stringent requirements, and the rest of these are D-T.
General Fusion uses a vat of molten lead and lithium, spinning to open a channel down the middle. A plasma ball is shot in from each end, then 200 steam-driven pistons slam into the container, making an acoustic shock wave that compresses the D-T plasma. The neutrons hit the lithium for tritium, heat the lead, and it all drives the steam turbine, with some steam diverted to drive the pistons. Jeff Bezos is a big investor in this one. Right now they've got the pistons working with the tight timing they need, and a prototype with a dozen pistons or so.
NIF says they're farthest along the development path, they're leveraging big advances in lasers, and with their LIFE development they're already working with thirty or so commercial suppliers to put together a practical reactor design with as much off-the-shelf equipment as possible. They don't need as much tritium inventory to start up as tokamaks, and think they can be competitive with fossil fuels. They've been talking 2020 or so for a demo reactor, but they also thought they'd manage ignition this year, and a recent DOE report said it'll be another couple years.
Helion targets a middle ground between tokamaks and NIF, in terms of temperature and density, which they think will be easier. They need $20 million for a full-size reactor, already built a 1/3-scale test unit. They're also working with NASA on a fusion rocket design.
Tokamak (like ITER) seems to be on a solid path to net power, but it's such a huge, complicated, expensive device that it's hard to imagine it being economical, and they're talking about a thirty year timeframe at best.
Here's the video of his excellent (and, in retrospect, sad) presentation at Google. I say "sad" because it shows an intellectual giant, nearly on his deathbed, practically begging an Internet giant, for what was really the tiny amount he needed to continue with what had become his life's work. Anyway, it's an intense, fast-paced, insightful, and (likely) final public look into his mind, his work, and his character.
http://youtube.com/watch?v=FhL5VO2NStU (Should Google Go Nuclear?)
http://youtube.com/watch?v=rk6z1vP4Eo8 (archive copy)
UPDATE: I found a good paper covering the above talk. It's a 25-page PDF, of less than 2MB, with color photos and charts from the video.
http://askmar.com/ConferenceNotes/Should%20Google%20Go%20Nuc...
I also got the impression that he was in his own world. That tends to be how geniuses talk. There was no introduction, no covering of the basics, no time to settle in. It was like jumping into the ocean of his mind, and you either sank or swam with his thoughts.
He was 78, and possibly already sick with cancer at that time, yet he came off as a young visionary with unwavering optimism. That, combined with his accomplishments, made him amazing, and one of my heroes.
He was asking for funding. For funding, you need to convince people that it's worthwhile. There are many ways to do that. One is to say "trust me", another is to persuade through education, a third is to use money which was promised for another purpose, and so on.
He used the first approach. Only, he did it in a way that's similar to what a lot of kooks do - throw graphs up on a screen, without the time to review them, and on a topic which few in the audience will understand. That is not a way to convince people to fund your project.
You say "That tends to be how geniuses talk." That is incorrect. Look at Fermi, Feynman, Gould, Sagan - all considered geniuses, and all renowned for their ability to explain things. Darwin's "The Origin of Species" was meant for a wide audience, and was not a technical monograph. Freeman Dyson has written some marvelous essays, including his "A New Newton" book review of Gleick's Newton biography. Just take a look at the Nobel Prize lectures and you'll see good evidence that Nobel Prize winners are also able to explain their work using something other than a blizzard of viewgraphs. Those people are ones I admire.
Instead, I think this sort of presentation tends to be the way that people who are convinced of their genius-ness talk. And again, it's the sort of way kooks talk.
I want to be clear here. I'm not saying that he's a kook. The polywell reactor may be the power source of the future. But that presentation detracted from his goal, I presume, of getting funding for the project. Really, if you didn't know it was Bussard, would you be convinced? If I had given the same talk, in the same style, would you be willing to contribute $10 million in funding?
I mainly backed the project because I find it fascinating that someone is doing this on a shoestring budget, 'competing' with other multi-million projects. The audacity of this is inspiring, although I have no idea whether the project will ever result in something tangible.
[1] http://www.kickstarter.com/projects/1992078142/building-the-...
[2] http://prometheusfusionperfection.com/
(edited for typos)