The next generation of aggregators are going to have a lot of interesting features, whether they are the ones I want or not. In reading John Tropea's Authority in your RSS readers, I realized that the aggregator needs to extend beyond the box of a single tool. I can see them reaching out to other services and pull that information in to the display of articles. I suspect the web-based aggregators are in a better position to take advantage of this than are the desktop based ones.
To use John's authority example, I could see an aggregator reaching out to a blog search tool to ask for the number of links to a given article. Or, possibly more interesting, the estimated total number of links in a thread of articles. To account for the fact that most blogs are low volume, I might be interested in knowing that a given post has generated more discussion than usual in the blogosphere. And this is made more complicated by the nature of online conversations: they develop over time, so I might not know that a given article is going to become important until later.
John also talks about slurping Waypath Related data into your reading habits, instead of just into your blog. In my threading view I get commonly-linked articles, I could imagine "related articles" threads too. Waypath has plug-ins for several blogging tools that let you add a list of "related" articles to the end of your articles. But this is for each individual blogger, not for an entire feed. Is is possible to get the data back from a Waypath quickly enough to add these related articles to your reading process? Waypath does have blog search capability that I use, but I don't know that they could accept the kind of rapid-fire queries that this would require.
People are also tagging their posts when they write, and bookmarking services let other people tag those posts. Why not aggregate/thread that information into the reading experience? What about geographic information?
With all this information, the reader is going to have to be smart about presenting potentially TONS of related content. One of the things I'd like to see SharpReader be smart about is the results I get from blog search engines. If it knew Technorati, PubSub, Waypath, Feedster, IceRocket, Google, etc. are search tools, it could probably collapse the similar items, so I don't have to see them repeatedly.
I am officially getting carried away. All this comes at a cost (at least today). The main reading experience needs to work, and work well. All the other features need to be easily accessed, but they can't get in the way of my normal process.