WebGuerrilla - 4:10 pm on Oct 1, 2004 (gmt 0)

This makes me think that Google is having trouble keeping track of large numbers of similar pages.

I actually think it is the exact opposite. Google has had the ability to identify and filter sites that are replicating content from a common source for a long time. That's why you don't see many ODP clones, or duplicate news stories from the major news wires showing up in SERPS.

For the most part, I think the majority of traffic/serp loss people are seeing is a result of them simply expanding that technology to also cover feed driven and serp generating sites.

Having a bit of experience with these types of sites, I can tell you that it is a very easy thing to do. There are so many common factors shared across these sites that it becomes quite easy to identify and filter them.

The trick to getting lstings restored is identifying the common items that trigger such filters and then stop serving pages to G that contain those items. (Yes, I'm talking about cloaking).