Comment by wilg

Perhaps someone can publish a paper to arXiv that provides a meta-analysis. But still there doesn't seem to be a clear reason to justify it, given that almost all internet text is not justified.