I agree with what most people have been saying here. I always base my vote on how well the review is written, how much detail it contains about the toy, and how "real" the reviewer was able to make the toy in my mind's eye. If I can read a review, and immediately understand the toys function/structure/usa ge/strengths/weaknesse s/details/etc, and imagine how well it would work for myself and my wife knowing our specific bodies/preferences/lik es and dislikes/etc, and doesn't leave me with any questions about it at the end, I give it an excellent vote. If it leaves me with one or two questions that I would like more detail on, I still usually give it a useful-extremely useful vote.
The only times I give really poor votes, are when a review gets posted that has not had any time or thought put into it, such as reviews where the writer is unable to convey to me how a toy works, or doesn't discuss WHY they liked/disliked certain aspects of it.
I would never penalize a reviewer for writing a review on a toy they didn't enjoy. Instead I think it's important that they've shared what they perceive as downfalls, in a well-thought-out manner. They should be rewarded equally as much as a writer who was able to express their thoughts equally as well, for a toy that they DID enjoy.