While certain aspects certainly stand out more than others, some of the judgement aspects are a bit off because the conclusion can be drawn by just doing something different from the viewers expectations.
In most examples, there is a specific action to a known goal, but in HotS, there’s some variety to the ‘correct’ action. Sure, hitting some skill shots means someone might be able to aim, but it may not indicate they know how to tank or heal; there’s a lot of different ‘skills’ to check, and not many players do all of them.
In watching some high tier replays and vods, it’s not uncommon to see some pub players thinking the stream is a newb/troll/boosted account when the choice from one play is different from their own expectation.
And yes, a skilled eye can generally notice some approximation of a skill level based on pressure from skill shots, making rotations, getting camps, but some of that is also automated and doesn’t necessarily align with actual ‘skill’.
In the heat of a moment, a lot of impulse reaction can look to fault a misplay that might otherwise not be faulted when seen through a replay and either player can see the perspective and priority of the other.
Part of the issue of the game is that many of the players are ‘objectively’ bad at the game, but part of the issue is that so long as they have any and everything else to blame, then do consider areas they can do better; they’re generally the ones that are ‘good enough’ and then claim every game has troll/afk/whatev that keeps holding them back. And more often than not, some scrutiny indicate they are exaggerating. Whether it be the frequency of the ‘problem’, the severity of it, the actual impact at hand, etc etc.
Human judgement may not stop working, but there are few things worse than wrong first-impressions.