I looked at the source blog post and decided it was pretty shoddy. I kind of agree that a reasonable person would doubt the veracity of this post, not just for the reasons loeg mentioned, but because the presented "evidence" is a single cherry-pick from data with lots of inherent variance.

I think it's valid criticism to say that you should have taken a closer look before submitting, and concluded that it was too junky for HN. In which case the flagging is appropriate.