Interesting statistical method. It's always going to be hard to control for groups attracting very small numbers of votes, which are favoured with a quotient method like this. Essentially what I conclude from your data is:
-Clive Cristian and Bond are utter crap houses
-Creed, Lutens and L'Artisan are the greatest houses
Let me explain:
-"over-rated", to an extent, is a self-selecting description. In addition to the aformentioned 2 crap houses which are unequivocally over-rated because of i) ridiculous price and ii) outright plagiarism, "over-rated" is going to attract the big hitters more than the small guys (ie to be over-rated, you have to be rated by a lot of people in the first instance)
-Creed, Lutens and L'Artisan attract the most attention in positives and negatives. They are the most polarising. This, in my opinion, is an indication of these 3 greatness. Like any art form, polarising opinion and attracting attention is a sign of impact and relevance. Any publicity is good publicity.
FWIW I voted for Creed in the positive, and Lutens / L'Artisan in the negative, but I'm happy toaccept that they are the 3 greatest.