While I think most people expect Facebook traffic to continue to grow for a while yet, it should also be clear that their moneterization efforts are pretty still primitive.
For example, did you know it's possible to get > 50% cheaper advertising on Facebook by selecting CPM rather CPC when you build your ads?
Even ignoring that, in my experiments advertising on Facebook is still hugely cheaper than on Google (taking the cost of a customer from acquisition to fulfillment). In any mildly competitive market (like online selling) that difference doesn't seem likely to continue.
Anyway, my point is that I think that while "completely taking over the internet" may be crazy, there are plenty of ways for Facebook to easily grow their revenue by huge factors.
Facebook's basic play is to build a lot of value into their network by bringing as many people in as possible, make some pocket change in the process, then eventually figure out how to milk the surplus value which will certainly be a large amount of value. My skepticism lies mostly in the execution of that second step; I'm totally unconvinced there's an ad bonanza large enough to fund this. Maybe there is. I think somebody's going to actually have to sell something and it's not clear to me what will work on the promised scale. That or they're going to have to raise the barrier of entry and if they do that they'll trigger the open internet competition in a heartbeat, competition already coming together and will be ready to go when that happens. The act of squeezing the surplus value of the network is the exact same act that will create their decentralized open competition.
But hey, who knows. Sometimes you get MySpace, sometimes you get Google. You tend to only be able to definitely tell them apart in hindsight.
It's pretty easy. You can select your target audience by gender, location, age group, and interest(s), and then select if you want to pay per impression or per click.
It is a good system (just as easy to use as Google's) and it gives good results. I've tried it, and that's why I'm bullish on them.
(It's worth noting that advertising has funded newspapers for hundreds of years, television for 50 years, and Google for 10 years and is proven to work in those cases. I shouldn't have to say it, but apparently some people continue to doubt that advertising is a reasonable funding model at all, regardless of the specifics of Facebook)
Another way to look at it: Do you think Yahoo or Facebook is more relevant to the average user, and what is the trend for each one, and what is the likely trend for advertising spend on each platform? (Note that Yahoo is currently worth $22 billion)
Basically, I think there expectations aren't as unreasonable as the large numbers seem to imply.
I want to jump up and down and yell "bubble" as much as the next guy, but I do have to admit this one seems relatively focused... at least at the moment. Just a small handful of companies trading way above what they "should" from a buy-and-hold perspective.
Some possible risks:
* Orkut-like sites were around for a long time, now nobody uses them today. People shifted to facebook and may move elsewhere.
* It is possible that a decentralized, autonomous solution may come up for social profiles.
* The fact that the company is not mature is a double-edged sword, a new idea or feature may come up that may well make facebook a less attractive place to be.
I'm not sure I can buy that, at least not at that price.
(Or, more accurately, a bet that enough other people will think so for long enough that you can unload your stock on them for a profit. I wouldn't buy and hold, I don't think.)