Also, I think that in reality there isn't as much brand loyalty as people think. Three years ago pundits said the same thing about myspace.
My position is that facebook is vastly overvalued - Gartner was just an example to show that even if they become the biggest in the field the income would still not justify their evaluation.
In fact Gartner has a market capitalization of 1.3 billion US$, meaning that facebook should be worth more than 12 times gartner. And Gartner has 4000 employees, 60.000 clients, and offices worldwide.
I think that is the key issue. I'll drop Facebook like a rock if it becomes annoying. That's generally true of most websites: they can only be "sticky" if they are a pleasant experience.
I think I've mentioned this before here, but (for example) Friendfeed is a nice comparison: it lets people show exactly where and why and how it draws links together. Facebook puts all of this information into the same facebook bubble.
No, I'm not a fangirl of facebook. I've logged into my fb account maybe a couple of times in a dozen or so months. Sorry. You want to assault my karma for this? Puh-leeze.
I look at structures: overall structures and value-added structures. Facebook really has none. If you disagree, please tell me why and how you disagree this setup is theoretically "profitable" or more profitable than, say, a random Internet page or a blogger blog. :)
I'm not sure any of Facebook's competitors have the number of users, or the kind of brand loyalty, that Facebook can offer. It's also easier for Facebook to be a social network that does market research (and more..?) than for Gartner to be a market research company that does social networks. Social networks are just a great foundation for a lot of other businesses.