Sci-fi author Charlie Stross had a wonderful take-down of He3[1] a decade ago. It's alas not very technical. The short of it is that there are no reactors, if we do make them they are going to be insanely fantastically hot (a billion degrees), there's only ~10 parts per billion (according to your wikipedia link) on the moon & mining that will be hard, and then we'd have to transport the stuff too.

I'd read something else about the challenges of reactor design, but not sure what it was. I haven't been enthused about He3 since. I feel like the one interesting use would be a Project Orion / Nuclear Pulse Propulsion system, but wow, boy does that seem like an expensive way to do propulsion: what's the use case?

[1] http://www.antipope.org/charlie/blog-static/2010/08/moonshin...