It's the opposite of the point you want to make. This is obvious if you simplify the example to something like flipping a biased coin. If the coin will come up heads 50-eps% of the time, and tails 50+eps% of the time, then the correct prediction is tails. Tails will often be the wrong prediction, but is still the correct prediction to make.
The outcome of this one event simply doesn't imply anything about the correctness of AlotOfReading's model. The model might actually be mistaken, but that's not an argument you've made.
Yes, that was exactly the point of my question. I was subtly accusing you of the same thing you are now overtly accusing me of doing. I was showing that we shouldn't look at everything through the lens of a probabilistic model.
The authenticity of the video was not a coin flip in which the results would be random within some sort of probabilistic distribution. We were not predicting a future event. We were observing and analyzing a past event. AlotOfReading didn't just come to the wrong conclusion. Their analysis of the evidence was flawed from the start and that is why I disagreed with them in that original thread before either of us knew the truth.
The outcome of this one event simply doesn't imply anything about the correctness of AlotOfReading's model. The model might actually be mistaken, but that's not an argument you've made.