Ed Batista in a comment to his own post:
"As users come to realize the value of their attention data, they're going to want more control over that data."
I agree with Ed - I don't think this is just an issue that only the 'cognoscenti' will be interested in. Let me explain...
I'm interpreting the word 'value' here not as monetary value, but the kind of value one receives from services using their data.
An example of the value I mean is Amazon recommendations. The value not only lies in Amazon's algorithms, but also lies within the data the algorithms use to provide the recommendations. A subset of that total dataset includes includes my clickstream, the books I've already bought and my wishlist. Without it, the recommendations cannot be done with the same effect. There is enormous value in this data to both Amazon and the customer. But it is locked-in.
The inability for customers to be able extract their attention data and make it portable for use on other services is the modern day equivalent of the business software 'lock-in' practices of the 80's and 90's. Businesses realized ten years ago that buying software that only uses propriety data formats that can't be accessed and re-used in other vendors' applications was a bad idea. They therefore demanded alignment with open standards from their vendors and ways for them to access their data for use in other systems. Now, 'openness' is a key selling point of most business software.
The same transformation will occur for the 'consumer' market. 'Consumers' will demand that service providers allow the extraction and portability of their data for use by other services, because they won't want to spend time re-entering their data across multiple services.
Consumers will want to avoid the attention data lock-in problem, just as businesses do today with their business data. The APIs that we're seeing more and more from online services is a step in this direction, but we also need non-API'd interfaces so the non-technical customers can get at their data. We need to make the data extraction process much, much simpler. Library Thing is a good example. I've invested time in entering data into Library Thing because I know I can get it out if I want to (as a CSV file).
Are the books I own and the associated tags I've created 'attention data'? Who cares! I just want to be able to get at my data. Why? Because there may be another service that I'd like to try out that provides me with value based on my data. And if I try it and use it, I want to be able to get data out again. My hunch is Library Thing has recruited more customers because the customers know they can get their data out again - its data openness is a selling point.
I don't think it is just the 'techies' that care about this. I admit to not having data on the following, but I'd guess that the ability to make customers' data portable drives use of a service. My bank allows me to export my statements, email services allow POP mail retrieval and contacts exports from their service, and then there are playlist exports, etc. It's hard say how / if these 'features' really drive customer sign-ups, but presumably they exist and have become standard features because there is a demand for them.
If I asked the non-techie whether they think it is important for them to be able manage and store their banking data, emails, contacts and playlists in a way that alllows them to use their data within any application / service they want, I would guess a large proportion would answer 'yes' - because they see the value in their data.