The recent Glassdog post about what's wrong with feed readers has generated some buzz, and in general I agree with this post. Feed readers need to do more - much more - to help locate and navigate related information, and this is something I've thought about for quite some time.
Before coding FeedDemon 1.0, I sat down and really studied the other RSS readers, looking for what was missing. It was (almost) immediately obvious that other feed readers did nothing to find information for you, instead acting as though you'd want to read everything. It seemed to me that as feed readers caught on, it wouldn't be long until people were faced with feed overload - too much data, most of which they don't care about. This was why my design specs for the very first release of FeedDemon focused on the importance of features such as "watches" and "search channels" which help you find information you're actually interested in. I also made sure to add basic filtering, grouping, etc., so you wouldn't be stuck with "sorting items by date," as this post suggests.
I considered doing more along these lines in 1.0, but I decided that since most people consider RSS to be brand new, they would need to be introduced to reading feeds using a familiar, email-like UI that wasn't overloaded with RSS-specific features. I know that sounds odd, but the more unfamiliar features a product has, the higher the learning curve - and the slower the adoption. So, my goal with FeedDemon 1.0 was to make it easy for those new to RSS to start reading, without the added complexity suggested by this post. Now that more people are using FeedDemon and other feed readers, of course, I can take things much further in future versions :)
One big problem I see with this post (and others like it) is that they talk about rating/ranking individual items, but I believe that the vast majority of users don't want to spend time doing this. Relationships between items can be made without requiring the user to do anything, so this will likely be my focus. That doesn't mean, though, that I'm ruling out ranking items in FeedDemon, just that I'm not planning to make ranking items central to using FeedDemon.
Jason Kottke ponders whether Google is developing a Mozilla-based browser. It's obvious Google is up to something more than search, as illustrated by its recent acquisition of Picasa. Kottke believes that Google is building an operating system, and I share this belief when I think about an OS that lives on the net rather than everyone's desktop.
Update: Steve Gillmor writes along similar lines.