diberry - 5:05 am on Jun 6, 2012 (gmt 0)
How are we defining false positives? And for that matter, false negatives?
It's been confirmed by a skeptical third party with SEO experience that my site was a false positive: it's clear to a human that I'm writing (however badly) for visitors, but a couple of things I did (like including a LOT of editorial links in some pages) could look to an algo like an aggressive SEO tactic.
The very fact that Google created a special Penguin reinclusion form, after not doing that in any other update I can remember, suggests they were expecting more false positives this time than in past updates. It may be that what Penguin's meant to do is right on the edge of being too ambitious: you need context to distinguish SEO from plain ol' marketing, or to tell whether something that technically looks spammy to a bot is actually just what searchers are looking for in a particular case. Those are just two examples.