That's a good question.
Obviously they are a reflection of perceived value. I guess the real question is do they *also*, in addition to reflecting the perceived value, help readers find the best solutions.
I think that rating systems help with the initial filtering, especially of negative value. Meaning, if an item gets any negative rating, it's likely to be actually less of a value (or - controversial!).
However, if an item does not get a positive rating, I don't think it necessarily means that it less of value. It might mean that the rating system isn't accessible enough, or that not enough people viewed this item, or a number of other things.
An interesting alternative is to rely on implicit information (such as the clickstream) instead of on explicit information (e.g. vote). Having discussed this with other folks before, it seems like this approach has serious downsides as well.
