I understand your analogy, but you lumped it into an example of "10 accidents". That analogy doesn't make sense. Collins made one prediction in 2020, and you concluded that he should never make a prediction again.
To your last point, nearly everyone else looked at what's being discussed and is drawing the conclusion that the context of Collins' single prediction matters. You did the same, allegedly, and arrived at an extreme conclusion on the other end. Nothing wrong with that necessarily, but the accompanying delivery has been suspect and yea, as a result it'll cause others to not trust a specific stranger on the internet.