Twitter was on track for 200m users in 2011 [1]. Twitter was valued at 7.7b dollars a few weeks back [2]. That's about $38.5 per user. Surely fewer than half the users would pay for Twitter, which would mean that those paying must be worth at least 77 dollars. At a discount rate of 15%, each of those paying customers must yield 11.55 annually in free cash flow (i.e. after costs) to justify Twitter's valuation.
Of course, Twitter could always find ways to extract money from the non-payers, but yes, you bring up a great point: Is it reasonable to expect that half the users will pay about $12 bucks a year on top of his fractional costs to continue to use the service?
[1] http://www.tgdaily.com/software-brief/52284-twitter-on-pace-... [2] http://www.mediabistro.com/alltwitter/twitter-7-7-billion_b6...
It certainly is not revenue ("i.e. after costs"), and is on a flow.
Believing that I said Rev(t) = Market Cap(t) is very wrong because I said Market Cap(t) * 0.15 = FCF.
(I think Twitter "pro" accounts are a bad business move for other reasons, in the same sense as Google Search Appliances might be more trouble for G than they're worth).
There are a lot of companies with freemium models that work, but they mostly treat the free version as a loss leader. I don't think Twitter is valuing its free accounts in loss leader territory. That's my take, anyway.