As an unscientific experiment, compare the rate of spelling and grammar errors in the comments on this article to the error rate in HN comments overall. I think I see a pattern, but it may be that all I'm seeing is my own biases.