TripAdvisor is just another tool in the research tool belt. We use it to help identify potential hotels we'd consider booking (not just on quality but also to ballpark pricing) but we'll always try to get supplemental reviews from other sources. (Or vice-versa, using it to supplement reviews from other sources.) The gem source of reviews are personal blogs IMO.

Personally, I read most TA reviews with a grain of salt. [Same thing with restaurant reviews.] The ratings are good for an initial parsing but one should get their key info in the actual reviews. And even if a review is not a fake, some reviews are just crap. Posters may have unrealistic expectations, don't provide details or examples to support their review, etc, etc. How much weight to you give a review from a one post wonder or to a newb?

My reviews are generally very long and as such it takes me forever to eventually write up my review from my notes. Normally I provide factual info (freebies, time it takes to walk to the train station, etc) and subjective info about what I liked, didn't like, and most importantly why, by providing examples and context so people can make up their own mind. Why I might like or dislike something may not have the same effect or impact to others.