Ok so a good sample size of a population is 10%.
This is absolutely bollocks. There are NATIONAL-scale surveys that are made on sample of 1000 units and are generally considered reliable.
If anything a legitimate objection would be that is a self-selected group. Incidentally, the criterium of self-selection here is "People who are already interest in the game" so not exactly a compromising trait for what the survey is asking about.
If anything that subreddit has a strong pro-Larian bias and tends to be insanely passive-aggressive toward any hint of criticism to the game.
You'd feel right at home, by the way.