It's worth pointing out where Gladwell is wrong about overnight successes and why.

While it's true that in a field where continued effort is required for continued success (like acting, to continue the theme of his 'American Idol' remarks), software (web software in particular) can be a different kind of beast. For sites that derive value from the network effect, your work can taper off once the network size reaches a certain point. (Witness Markus Frind of PlentyOfFish.com.)

Furthermore, while times are tight now, the 'get bought' method of internet riches, where/when available, has to qualify as an 'overnight' success.*

*Not to say that this is the most likely or desirable case, just that it qualifies as a counterexample.