Just remember one thing with all this ongoing discussion: we don't know who the people are who are taking the triangle test, and we don't know what they've been eating and drinking prior to the test.
Both of those issues seriously call into question the validity of the results. No amount of statistical handwaving can overcome a lack of good data to begin with.
If people have been drinking IPAs prior to testing a lighter more nuanced beer, can they do that? If this wasn't an issue, then why do people suggest, when sampling a flight of beers, that drinkers move from the lightest to the heaviest beers?
Of course it's an issue, and it's the biggest flaw in how the brulosophy experiments are done. We don't know the population to whom the sample generalizes, and we don't know how or even if they were prepared for the test (clean palates, etc.).