I disagree, and think you have to use every data point available. While I agree that many/most reviews on social media in the first weeks will be from non-riders, a quick read of the Google reviews shows some people who seem to have ridden it (to the extent we can know). Every set of YouTube comments, forum responses, Google Maps or Yelp reviews, etc. is a data points. The data is likely biased and comes from a non-representative population of parkgoers-at-large, but the best we can do is take these data as they are because they obviously represent something.
e.g. the fact that F&F is majority 1* reviews (while Gringotts is majority 5* reviews) can tell you something, even if it's not "F&F is 5x worse than Gringotts", or if F&F has 10x more 1* than 5* reviews, we won't assume that "For every person who rode F&F and loved it, there are 10x as many people who hated it", because haters are probably more likely to go to social media to complain.
But I think it's interesting, and it's very, very relevant, as these parks all have a heavy social media presence and social media now constitutes a massive component of the advertising industry.