I used to have high hopes for fusion power until I came to learn two things:
1. Neutrons rapidly destroy any kind of containment. He-3 is seen as some kind of holy grail here. Unfortunately it's incredibly rare. The best source is probably the Moon (which, with no atmosphere, has had billions of years of solar winds to collect non-trivial amounts of He-3); and
2. To sustain a fission reaction any current research reactor I can recall reading about uses probably radioactive isotopes (most likely tritium ie H-3) or at least somewhat uncommon isotopes (ie deuterium, H-2).
Fun fact: per cubic foot compost generates more power than an equivalent volume of material from the Sun's core [1]. What makes stars such great energy producers isn't fusion per se, it's their sheer size and fuel supply.
Fusion goes hand-in-hand with the Utopian concept of "free energy" because obviously water (and thus hydrogen) is abundant. But no energy will truly be "free" although it may be cheap enough to have the almost the same impact as being free.
There are four component costs in energy:
1. The cost of building a device that produces energy;
2. The cost of maintaining that device;
3. The R&D required to bringing that device to market. Obviously, over time, this cost diminishes to zero; and
4. The cost of extracting, storing and transporting whatever fuel is required.
So for hydrogen, assuming a fusion reactor doesn't use sea water, you must first extract hydrogen. This is actually relatively expensive. Hydrogen extract is seen as a way of smoothing out power generation from renewable sources (most often wind power) [2].
Sure a fusion power installation could produce the energy for this but remember that any power system needs to produce more power than it uses or its largely worthless (barring corner cases like producing energy in portable form, which is the entire model for batteries). With hydrogen extraction, you've just raised the bar on how much power than reactor needs to produce.
Likewise, hydrogen storage is non-trivial. It's flammable and hard and expensive to store in liquid form (compared to, say, liquid nitrogen [3].
Energy needs to be produced in many forms. Fusion is on the scale required for power stations but that's only one of our needs. What about vehicles? It's really hard to beat the amount of energy stored in oil (by mass or by volume) and the ease with which it is extracted.
I see a long term solution to this in genetically engineering microbes to produce readily consumable compounds for abundant and sustainable inputs. Another fun fact: if you've read Neal Stephenson's Anathem you'll probably recall his mention of "fuel trees" that suggest this kind of approach.
In the far distant future I see the only viable power source that would allow us to live in space, let alone cross interstellar space, to be black holes [4].
[1]: http://en.wikipedia.org/wiki/Solar_core
[2]: http://en.wikipedia.org/wiki/Wind_hybrid_power_systems
[3]: http://hypertextbook.com/facts/2007/KarenFan.shtml
[4]: http://io9.com/5391989/a-black-hole-engine-that-could-power-...
