And to the user who first sees the pay-wall, you're asking them to make a non-micro-transaction commitment. Maybe the page view only costs two cents, but they've got to create and charge some sort of account for, say, $10. So they never sees that "It's two cents!" pitch. Their eyes lock on that $10, they laugh/curse, and go get the content elsewhere.
Unless consumers are ready to spend all $10 that-visit (as at a GameWorks or similar) the buy-in on that sort of 'charged account' just doesn't happen.
There's a patent and everything. http://www.patentstorm.us/patents/7110979.html
I think Stephen Glassman (formerly at DEC) is now at Google. Not sure if he is/was involved in Google Checkout.
Flattr is the most interesting approach to this "problem" yet.
Even if I'd be willing to throw a few cents at a site for content, I don't want to log in to yet another service with yet another username and password to do so; I don't want to click through three pages to authorize a payment; I don't want to do much of anything apart from perhaps clicking "Yes." Most importantly, I don't want to have to create a new account with some random micropayment company that's probably going to be gone in a few months in order to endure any of the other annoyances listed above.
This is pretty much the only reason Apple is getting away with selling stuff on the iPhone - you have to be associated with an iTunes account to use the phone, and all you need to do is punch in your password when it comes time to purchase. Since the same amount of annoyance comes at you whether you download free or paid apps (a brilliant design decision, by the way - there's really no need for them to force you to put in your password to authorize a "purchase" of something that's free since they're not charging you anyways, but setting up that equivalence between free and paid apps is useful to them), you can actually decide between the two on the basis of price rather than annoyance.
I think it's much harder to conceptualize paying for a news article as providing some sort of ownership interest. Fact is, there are several authors and essays I read over and over again, but that's probably less than 1% of my internet reading. You're basically asking me to pay to own something I don't want to own.
Granted, news orgs are spinning micropayments as paying for a service, but each individual article fits more naturally into my idea of product (tangible, doesn't really change) than service.
You'd think Paypal, or Google, or someone would have enough clout to work out some deal with the credit card companies so that micropayments could be dealt with through pure percent-of-transaction fees (relatively high ones, even) instead of having fixed floors on the transaction costs...
The problem is not so much the computational resources. It's the other resources, especially the human resources required to deal with exceptional conditions. One common exceptional condition is fraud: Either the buyers will try to cheat you, or the vendors will try to cheat you, or certain buyers and vendors will get together and conspire to cheat you -- by, say, laundering money through your operation, which will be a problem for you when the IRS and the FBI come to audit your books.
Every transaction has a failure rate, and with such tiny transactions it takes a very low failure rate to destroy your margins. If one out of every ten thousand 1 cent transactions is fraudulent, or gets disputed and charged back by the customer, or fails verification, or sets off a fraud warning at someone's bank... and the resulting exception ends up costing more than $10 worth of time and resources to fix, that's 10 percent of the gross.
Here's another random fact: In the USA it now costs 44 cents to mail a first class letter weighing up to 1 ounce. That will typically allow you to mail 4 or 5 sheets of paper. Which means that it costs of order 10 cents just to mail a sheet of paper. One of my double-sided credit card bills seems to hold about 40 transactions, which implies that it costs about one-quarter of a cent just to mail the paper report of a single transaction. If the average transaction is for a couple of cents... the mailing costs to the credit card company are somewhere around 12.5% of the gross.
That's just one random example of a cost that doesn't scale uniformly downwards as transaction size goes down. There are many others. The cost of a transaction just isn't proportional to its size.
Obviously you can mark up a transaction enough to make it worthwhile. But if that markup is very high, your 2-cent-per-article micropayments operation will end up with a huge competitive disadvantage against someone who collects up articles into bundles and then sells the bundles for $1.99 apiece -- or, better yet, an annual payment of $23.99. We call those people publishers and editors.