> I don't think you're characterizing the results accurately, particularly point #2. I don't think they really had a way to measure #3. #1 is just obviously true but doesn't add much.

Maybe not, but #2 was from a blog that specifically looked at the data of messaging: how many men received messages, how many women received messages.

IIRC, nearly all the women received messages and sent messages, while only about 10% of the men received messages while all of the men sent messages. I don't really see any other way to interpret data showing that both men and women sent messages at about the same rate, but only about 10% of men received any messages.

Now I don't have the blog handy (and I rather wished I did), but my takeaway #3 was from reading all the blogs available, in basically one marathon sitting, not from one particular blog.

It strikes me that if women and men sent messages in exactly the same long-tail shape of distribution, and you had a knob that controlled how frequently each sent messages, you'd be able to turn each knob until 100% of women and 10% of men received messages.

This is not to say you're wrong or right, but I don't think the numbers tell us much without seeing the full distribution.