First, when zoomed out, outliers in all possible tasks become more common — internet commenting is just a subset for silly folks like me.
Secondly, the emergent human social fabric is built to recognize and amplify outspoken and / or talented outliers, via mechanisms whereby others who {agree, can find utility, can profit} are incentivized to act as amplifiers. The cost function to repeat a message drops precipitously every time it’s repeated (influences status quo). I’m not sure it’s particularly surprising that internet social forums behave by the same rules — and are even optimized to replicate them mechanistically (upvotes).
I mean... not be dismissive, I guess it does strike me as particularly neat that the internet provides a medium for these people to productively share insight and identify new niches where they can potentially add value to the rest of the world. Where would we on HN be without, say, patio11? :)
Let me try to explain with a bit of an overstatement: Most TV is crap, but year after year they keep making it. People making it cannot be good at it? Well actually they are. They found the sweet-spot by maximizing the profit in terms of eyeball captured they will make from the least amount of effort. That is success.
Now-a-day successful politicians are far better at making people vote for them than actually realizing the platform they are elected on. They are literally good at the game of democracy, but don't know what to do with the spoils. The difference between those two seems to be "fake-news".
Lets assume that the prolific reviewer on Amazon is completely legit. He is obviously good in the sense of efficient at reading and writing reviews. That we do not see the "good" in an outcome of having so many reviews written by the same person does not make his activity less good as an activity.