This was my reaction to the article also - the base assumption that the relatively tiny group of audience members who rate films on imdb are statistically representative of the audience at large needs to be proven first.