Maybe it isn't the volume of talk, but the quality. Maybe the stereotype comes about because women talk about shallow topics like shopping and clothes, while men talk about important stuff like football and motorcycles.

Interesting, but in my opinion the research presented (in the article, at least) doesn't answer the most pertinent questions. It doesn't surprise me that much that in male-female conversations, the sexes talk about the same amount, but what about when men interact with men and women interact with women? And even more importantly, what is the content of men's and women's speech--not just the volume? The prevailing wisdom is that men talk mostly to convey information, whereas for women, emotional interaction is just as important as information exchange.

It has been my experience that women are less direct. An example: Several years ago I said to my mother (we live together, she's now 102) that we hadn't had a dog in a long time and I wanted to get one. Her response was "you won't feed and take care of it". What that really meant came out several days later when after thinking to myself that I would feed and take care of it, I brought home a puppy. Her response to my walking in with the puppy in my arms was "that dog is going right back where you got it tomorrow" which shows what she really meant in the first response. I have talked this idea over with male friends one of whom worked as a nurse thus working with many women including superiors and we all agreed that it was a valid observation. By the way we still have the dog and she's a great companion who my mother loves dearly, but that's another story.

Two thoughts on Mike O's comment:
(i) This kind of indirect mode of communication is exactly what one would expect from speakers used to thinking of themselves as lower-status speakers, in accord with the claims of the main post.
(ii) You should, as a general rule, more or less completely discount any observations you make of the sort reported here, i.e., anecdotes backed up by others' reports of similar anecdotes. Your initial hypothesis about such putative generalizations is that they are spurious, and the result of confirmation bias. This isn't at all particular to gender-related generalizations. It's just that this form of over-generalization is something that our species is very prone towards. That's what the practices of science are for, in large part -- to help us overcome our own innate epistemic deficiencies.

The wikipedia entry on confirmation bias is pretty good, btw:
http://en.wikipedia.org/wiki/Confirmation_bias