I think this is fascinating: GalaxyZoo.org is asking volunteers to look at pictures of the sky and scientifically classify entire galaxies. Spiral? Elliptical? Clockwise? Anticlockwise? It turns out the human eye and the human brain are much better at this sort of thing than any currently available technology.
But isn't science about precision? But what if they're wrong? My bet is that the underlying system relies on the establishment of a consensus.
Most people are reasonable and of good will, which is why free societies work as well as they do. If you design a system that accepts the fact that some people will make mistakes, and others may be vandals, but you allow the larger number to define the outcome, then you accommodate the realities of human nature. This is the principle behind open processes ranging from Wikipedia to democracy, and it's something that we in the media business need to learn to use.