I'd say probably closer to 20-25 years before we see a controlled thermonuclear reaction used as a power source. There's a promising experimental design that's scheduled to begin production soon (ITER, will be built in either France or Japan)-- although construction could take up to 10-15 years and I assume that it would be another 10 years at least before we had a real working model.
Well the last was sarcasm, but for real we'll never see this power source being used for electrical power and such. Too many jobs and money at stake. There is an unwritten rule that goes like this: If you invent something that will significantly impact the world's economy, then it will either be squashed, bought out, or forced to move at a slow pace as to be unprofitable (hydrogen fuel cells).
What we really need to speed things up is a worthy adversary(s) and eminent demise! Near death concentrates the mind wonderfully.
Fission power went from first controlled chain reaction (Fermi's pile at Univ of Chicago in late '42) to viable controlled power production in the prototype submarine reactor at the Idaho National engineering lab in early 1953. That's just slightly over 10 years from first 'proof of concept' to useful device.
Ah well, maybe when oil hits $1,000/bbl we'll feel a bit more pressure to get off out collective duffs and get serious.