I got a hold of the research article that is linked from the main article. It's a meta-analysis of research on the topic. It was done in 1999 however, and the research it's based on is likely considerably older than that. I'm going to try to read through it.
I'd be interested to see what recent research says on the topic. I do know that cognitive research in habits and productivity talks about the importance of rewarding stretches of work with breaks, fun, and other rewards, and that would seem to conflict directly with the intent of the article.
I'm very much on the fence with this one. I love the idea of intrinsic motivation and totally agree with Rob, but I've also found extrinsic motivation in many forms to be helpful in teaching, learning, and my own productivity.
One example where I've seen extrinsic motivation work (at least it sure seems like it works to me) is with Wayzata's Middle School Boys Cross-Country program. They usually have tons of kids who are new to distance running, and at the beginning of the season that offer a "10-mile club" T-shirt to anyone who can complete a 10-mile run before the end of the season.
For kids who haven't run more than 2 miles or so, that's a seemingly impossible goal, but a ton of kids get those T-shirts. Our son still talks about how driven he was to get the T-shirt by the end of the season. He honestly didn't care about the item itself, but it served as a tangible icon for an excellent intangible BHAG (big hairy audacious goal), and he still talks about how it motivated him to go to practice and work hard. He now plays soccer all-year round, and he attributes the conditioning he got from running cross-country in middle school as an element in why he's successful in soccer. (It's also worth noting that the school's cross-country program wins national championships and is one of the strongest in the state year in and year out.)
I've seen lots of cases like this, where extrinsic rewards are used in ways that I can only say are effective. And that's why I kind of lean back on the hunch that it's not as simple as "extrinsic motivation is bad." It seems like a much more complex concept, with implementations that are weak and ineffective, and implementations that are beneficial and effective.