There is a distinction between "revenue is not important" and "if lots of users love you, you can probably figure out how to make money out of it (if you arn't acquired first)".
I think another difference is that there are companies absorbing the risk and associated losses. As was noted in another one of Paul's essays, big companies tend to be good at extracting revenue from an existing product but not so good on boldness and innovation.
There are a variety of business models to try once you've made something people want. Sometimes it is ads (myspace) sometimes it is merchandise (homestarrunner) or a combination of the two (penny-arcade). Maybe your content is worth a subscription (hotornot). Maybe you sell books or offer training and advanced services (SysInternals/Winternals).
Another argument for not worrying about the business model, if it is your intent to be acquired, is that your product may wind up more flexible. 3 different companies might use 3 different ways to make money from your product. Again, the distinction is that it's already profitable companies doing the buying, rather than the Stock Market At Large.
