Well, I like the polywell approach, and the main argument against it (electron leakage) has already been disproved in extensive testing last year. The U.S. Navy is funding it and is satisfied with the data but still hasn't coughed up the $200 million needed to build a full-scale prototype.
This kind of fusion scheme was conceived by the late Dr. Robert Bussard, a leading physicist in the field of fusion research, after whom our Trekiverse Bussard collectors were named. When he died, he left a team of competent physicists in charge, who are still working with the few million a year the Navy provides to keep the lab open and the project going but still not the $200 million it would take to build a full-scale working reactor, which would be about 5 feet in diameter. Instead, they have built one miniature after another too small to demonstrate overunity operation.
Basically, the idea is to have 12 magnetic fields squeeze the space in the center, and pressure can be increased just by ramping up the voltage. Originally, some scientists said it wouldn't work because of electron leakage to the outside, but the latest miniature prototype worked fine, with no leakage. The device is basically just 12 electromagnets in a dodecahedral configuration, all bearing down on the center, where fusion can occur if they use big-enough magnets. Even if they get the funding, there might be some unforeseen problem, but all the data gathered so far suggest that it will work if they get the funding and build it in the theoretically minimum size to produce overunity fusion. It's a gamble, but if it works, the ability to build such small and relatively cheap fusion reactors would obviously have an enormous positive impact on the world economy, not to mention the environment.
Another interesting aspect of this design is that if it works this kind of reactor can also be configured as a fusion rocket for spacecraft.