The only way reviews can be faulty is if they mess up objective facts/details/information; otherwise they can't be faulty because they are subjective views.
Reviews are a tool for measuring probabilities. If 80% of people like a given product, chances are higher that I too will find it worthwhile, even though I could easily end up in the 20% that don't like it.
Reviews aren't supposed to tell you how you might feel about an end product; they are simply there to tell you if the end product is worth trying/spending money on/investing your time in/etc. They're like commercials; you might like what you see in a Big Mac commercial, but you can only really know how you feel about it if you try one. But it nonetheless remains that commercials can be a useful tool in helping you decide whether you want to try a Big Mac or not.
For books, I seek to examine the actual premise/storyline, the author, and the critical and popular consensus (i.e. reviews), and weigh them against my tastes and experiences. These tools are what I use to decide what books I'd like to read, as it grants me a greater probability of finding books that I will like, as opposed to not.
Sounds like you have tended to trust reviews fully in the past before getting into something, and ended up being disappointed when the reviews did not match up with your own views on the end product -- thereby meaning that you have come to see reviews as an unreliable tool when it comes to informing you about what you should get into. Even so, if you have a system or approach that works for you, by all means pursue it (whether that course has reviews in it or not). 8D