diberry - 4:30 am on Apr 17, 2013 (gmt 0)
I don't believe (unless I've misread) anyone's arguing the teams are collaborating or conspiring. We are, however, saying that the most basic quality control requires that someone at Google look at the final result - a SERP page with organic and paid results - to make sure it's working the way Google feels it should.
The debate is over how "Google feels it should", and whether they would manipulate the algo (as opposed to blending the ads in better and so on) to encourage clicks. And you have to admit, they have a big margin for error so they could certainly experiment with such a thing without much risk.
But on the other hand, everything we're seeing in terms of "brand bias" could be explained as a shortcut for an algorithm that can't handle the monumental data being thrown at it, which is increasing exponentially all the time. Letting trust whittle down the sites to something manageable, if not perfectly relevant in every case. Again, that big margin of error (and the fact that no one can do it better) makes it plausible that they could do this without much risk.