"Nobody reads from just 15 or 20 sites a month. People read from hundreds of sites a month, creating a vast long tail of publishers."
And somewhere out there Ted Nelson is saying, "I told you so. We needed transclusion, transcopyright, and micropayments in our hypertext from the beginning!"
This is why I think that people that predict that custom native apps will displace the web are getting ahead of themselves. I'm willing to install a dedicated app for a handful of services I use all the time but otherwise I'm happier to just keep a bookmark.
The cost of my deciding whether or not an article is worth whatever the cost of reading it is, let alone the huge privacy invasion that per-article pricing implies (tying information content and payment to a single user) is a non-starter for all but a very few sites.
That leaves us with accompanying advertising, patronage, syndication, content-as-promotion (propaganda/advertising), and self-publishing.
"Accompanying advertising" is the existing model of display or other ads accompanying content. This works for some sites, but not particularly well, and there are concerns in the mobile space.
"Patronage" is essentially sponsorship of work, as fine art patronage was fairly common in the past.
"Syndication" would require a model of mechanical payments and essentially a tax-like funding stream (part of your monthly data plan) from which content providers are paid.
"Content-as-promotion" is where the initiative for creating content is its own utility in promoting the commercial or ideological goals of the author. Bands promoting live acts, political or religious organizations.
"Self-publishing" is essentially vanity publishing -- where many blogs are (possibly a special case of content as promotion).
Of this list, the syndication model is the only one I see as having strong viable economic potential for general works, though it requires the creation of a syndicate. Small-scale syndicates, in the form of paywall/paygate operators could be the start of this -- major periodical publishers (NY Times, Murdoch/Dow Jones), Apple, Amazon, eBay. But for general content, I'd see a mix of a mechanical reproduction right, a flat (or very simply tiered) level of content payments, and what would amount to an all-you-can-eat or bandwidth-based scheme through a data service provider or independent syndicate provider. It's a leaky system, but might just work.
What about Flattr model where there’s no mental transaction cost as everything is prices equally? (You have a monthly budget, one click payments, at the end of the month budget / number of clicks = click value)
Works the same way in mobile apps for example (although you need to OAuth your account initially) and you can keep your privacy so publisher only knows that someone paid, not who did so.
Economics and financial transactions work well with fairly evenly paired buyers and sellers, and largely undifferentiated goods (bulk items, water, food, etc.). Where market power is unbalanced, you end up with monopolistic or oligopolistic markets, where goods are highly differentiated (think: enterprise software sales) you end up with very prolonged negotiation over quality and deliverables (hence the high cost of such sales).
Information economies violate both precepts.
We've never had a market for unbundled information goods purchases. A newspaper or magazine is a bundle of articles, a book is a bundle of chapters, cable TV is a bundle of channels. Most entertainment parks sell a bundle of rides rather than tickets to spend on any given ride. Even music and video sales are a very different experience when you can sample and preview what you're getting.
If I had to make a purchase decision for every single thing I read, I'd go stark raving mad (OK, this may explain something).
Commercializing every single transaction is not feasible.
I don't think it's really been a "them" for a while now; Xanadu as a funded project with employees wound down around 1993-94, and since then it's just been Ted Nelson occasionally sketching out designs.
And somewhere out there Ted Nelson is saying, "I told you so. We needed transclusion, transcopyright, and micropayments in our hypertext from the beginning!"
http://en.wikipedia.org/wiki/Project_Xanadu (p.s. not a xanadu zealot, just fascinated by the history)