Having read Saint Sir Thomas More's Utopia some time ago, I can tell you that an "artificial leaf" isn't going to cut it. If you haven't read it, I highly recommend it. It's a short read and has a lot of interesting things to say.
Okay, onto the meat of the post.
I would first like to object to the nomenclature being used here. This is not really photosynthesis which uses light, water and carbon dioxide to make sugar and oxygen; this is photoelectrolytic splitting of water – it uses water and light to make hydrogen and oxygen.
And this is not going to create Utopia. The idea that this will "make each home its own power station," seems... not well thought out.
First of all, there is the water. A real leaf has whole piles of systems dedicated to collecting, purifying, reserving and transporting water. I'm going to guess this system needs pure distilled water – about 6L of it for a 20 kWh system. 6L of water may not sound like much to people living in developed countries but it may turn out to be difficult to obtain in areas where a device like this could do the most good. And if you have enough sunlight to split that entire 6L, you only get 20 kWh of energy, which is approximately the average daily energy consumption for Africa in 2008 so realistically you'd probably want more in order to continue to have electricity if you have a cloudy day or two.
The second barrier to entry is of course the fuel cell. The catalyst described in the press release uses light to turn water into hydrogen and oxygen; you then need a way to turn those into electricity. I don't know a lot about buying fuel cells but the Fuel Cell Store sells a 1 kW PEM fuel cell for $4000. That doesn't exactly sound like developing world household price to me.
And finally, there is the problem of storage. Hydrogen is not very dense. If you wanted to store that hydrogen unpressurized, you'd be looking at 7 and a half cubic meters of storage space. If you want to pressurize it, that's going to cost you some energy, making the whole system less efficient. And let's not forget the other obvious problem about storing hydrogen.
All that having been said, I think this is a useful scientific discovery – a stable, inexpensive catalytic surface for photoelectrolysis. Perhaps very small systems can be made cheaply that would provide a few hours of reading light for people in the developing world. And I can see this being used in the developed world if it is cost-competitive with traditional solar PV at a system level, that is, if the cost of the electrolysis unit and fuel cell less per kW than PV cells. But if past experience with commercializing research like this is any guide, we can expect at least 5 years, probably more like 10 or 15 before that is even a question anyone has to ask.
My complaint here is not with the discovery, but rather the way it is being reported. Science can be useful without leading to an instantaneous paradise. Not every discovery is the Holy Grail of something. In the "what have you done for me lately?" world of academic funding, it's hard to be humble, but I think in the end, by providing expectations to the public which can never be met by any single scientific discover, we're only making things worse.
Update:While following up on this a bit, I discovered a news article with enough technical detail that I was able to figure out which publication was being described. It turns out that it will, in fact, run on non-pure water.