Julian Sanchez, a frequently sighted figure in these parts, takes on Malcolm "Tipping Point" Gladwell in Ars Technica:
Gladwell's model, which has itself become something of a cultural epidemic, posits that a few hyperconnected "influentials" are the key to the runaway viral spread of fads, fashions, ideas, and behaviors.
These pivotal individuals, according to Gladwell, determine which trends will wither on the vine and which will "tip," becoming mass phenomena. But [Duncan] Watts, a pioneer in the mathematical modeling of social networks, has tested the "tipping point" hypothesis, both empirically and in computer simulations. As it turns out, according to Watts, it's just not true. There are exceptionally well-connected folks out there, but they're so swamped by ordinary individuals that they can't account for genuine cultural cascades, which result not primarily from the activity of social "hubs" kick-starting trends and broadcasting them to the masses, but average Joes and Janes passing them on to other average Joes and Janes.
The whole argument is here. Sanchez suggests that one reason Gladwell's theory is popular is because it flatters readers. "How tempting, if you're a marketer, an activist, or just the sort of person who tends to pick up books by New Yorker columnists, to imagine that you have it in your power to launch a runaway cultural phenomenon just by targeting a few key people. Why, you may even be one yourself!"