But as Sullivan would probably agree, the choice between a paywall or a tip jar is not as clear-cut as it sounds: realistically, it’s more of a spectrum. Some paywalls are forbiddingly high “Berliners”: if you don’t cough up, you have no access. Most, however, are porous to a greater or lesser extent. The Times and Sunday Times of London will give you the first 75 words or so of any story; the New York Times will allow you a certain number of free articles per month, plus all articles arrived at from external sites; the WSJ will let you in if you’re coming from Google, or from a link which has been emailed to you by a subscriber. At other sites, the wall is drawn around some content but not all: the New Yorker, for instance, puts only some of its magazine content online for free, while the Boston Globe hides all of its content behind a Berliner paywall but then allows a subset of that content onto Boston.com for free.

None of these models is obviously better than any of the others. No paywall lasts untouched for long: all publishers are making decisions to put up or take down paywalls every day, and it can be hard to keep track of which publications have which model. (Right now, for instance, without looking, I genuinely can’t remember whether the Economist is paywalling any of its content or not.) Just in the past few days, we’ve seen one high paywall demolished, at Variety; there, the new proprietor, Jay Penske, called it the “end of an error”. Meanwhile, another paywall has been erected, at Fortune: for the time being, for now, most Fortune magazine content is now behind a wall, while online-only content is free.

In an editor’s letter which isn’t online, Fortune’s Andy Serwer says that “we consider Fortune’s content valuable enough that we have decided not to give it all away online”. The unfortunate implication is that the online-only content, including the excellent Term Sheet blog, is not valuable enough to be worth charging for. On the other hand, if you look at the pricing, you’ll see that the cost of a digital subscription — $19.99 per year — is exactly the same as the cost of a digital subscription plus home delivery of the magazine. And the unfortunate implication of that is that all the extra value one finds in a magazine — the art direction, the layouts, the ability to read it while waiting for your airplane to take off — is also worthless. (Contrast that with the NYT paywall, which doesn’t really charge for the content at all, but rather for the online ability to navigate from one story to another.)

The real reason why Fortune put up a paywall, of course, has nothing to do with how valuable Andy Serwer thinks the magazine’s content is. Instead, the paywall is just another way for the Time Inc. brass to try to make money and keep the magazine’s rate base high, the idea being that people will be less likely to cancel their magazine subscriptions if they know that they won’t be able to read that content online for free.

Which brings up a fundamental rule of online subscriptions: there is zero correlation between value and price. There are lots of incredibly expensive stock-tipping newsletters which have a negative value: you’d be much better off if you didn’t subscribe to any of them at all. And of course there’s an almost infinite amount of wonderfully valuable content available online for free, starting with Wikipedia and moving on through the sites of organizations like Reuters, Bloomberg, the Guardian, and the BBC.

Or look what happened when Newsweek and Sullivan parted ways: both of them started subscription products, at almost identical prices. (Sullivan wants $20 per year; Newsweek wants $25.) That doesn’t mean the two products have almost-equal value; it just means that both Newsweek and Sullivan — just like Marco Arment, for that matter — came to the conclusion that the $20-a-year range was more or less the point on the supply-and-demand curve where they would maximize their income. They might be right about that; it’s hard to tell. Paywalls are put together in so many different ways, at so many different price points, that trying to work out their relative merits, in terms of income generated, is almost impossible.

Felix Salmon is an Audit contributor. He's also the finance blogger for Reuters; this post can also be found at Reuters.com.