I went to the Yahoo site explorer to see if it had picked up all the 301s of a site section to their new location. I was surprised and pleased that that had updated all the moved pages except one... that is, the old URLs are out of their index, and the new ones are in. All these pages were moved in the past month.
I was was additionally surprised to see four ancient URLs (four YEARS old or so) still sitting there in the index. The correct URLs are also in the index.
So, why would yahoo be able to promptly clean up dozens of redirects, but not want to clean up ones from 2004? For each of these ancient URLs Yahoo lists about ten links to them from domains not under my control. These links are usually a mix of PR1 to PR3.
And then more mysteriously, these four ancient URLs were part of a group of about a dozen moved at the same time in the exact same way... so eight are gone, but four continue to be shown four years later.
Assuming the redirects of the four still in the index are set up the same ways as the eight removed years ago and the dozens removed this month, why possible cause could lead to yahoo hanging onto these ancient URLs?
Almost certainly an artifact of the integration of the Ink d/base - it didn't go 100% smoothly, and there have been examples of data for sites getting messed up, such as URLs being handled incorrectly, as in this case. If you ask Y nicely, they might get someone to look at it, and manually reset the data.
If the URLs postdate the end of that incarnation of Slurp, I dunno - all SEs just flat out get things wrong occasionally, or it could possibly be indicative of old links casuing the URLs to live on in crawl control. If that's the case, ensuring they get correctly redirected could get you some long-lost link juice back