Would it be possible to use that algorithm and get the same sort of 95% confidence here given that many people can't downvote?
But green is not reliably "new". It tends to be strongly correlated to newly created accounts, but I've seen weird behavior (like a child comment is marked green... whereas the comment thread was started by the same commenter in gray in the first place), which leads me to believe there's something subtler going on. I thus have no idea what the difference in color is supposed to indicate at all.