This is a good move on Google's part. It will bring enhanced, richer results to all. The mentality at Google is improve, improve, improve. While certainly not perfect, I give them credit for a positive attitude and willingness to take risks.
The mentality of Google is collect, collect, collect. (All the information about you that they can get). They probably want to avoid another $22 million payout when they spied on Safari users, which was found because Google changed the code that allowed their hack in their own Webkit branch.
----------
Actually, it probably is as they say it is. Chrome's multiprocess model is vastly superior to WebKit2's model in terms of stability (while also being much more resource-hungry). Apple has jumped on-board with the pointless WebKit2 multiprocess model. I believe most of the code Google will be removing is the code support for this. That will be a much better engine for google. I suspect this fork will follow WebKit closely where it can though. Besides for the multiprocess model, WebKit is still an amazing engine.
Yeah, sandboxing to prevent exploits from doing damage to your machine is a "pointless multiprocess model".
Google currently has 95 webkit reviewers versus Apple's 59. And Google currently responsible for the majority of commits (49% vs Apple's 25%).
What's the size of these commits? If Google tries to impress people like you, the easiest thing is to order their developers to make many, many small commits instead of a few big ones.
----------
What personal data does Google sell to advertisers?
1. As Google fans love to point out, they don't _sell_ your data. They rent it out. So they can sell (sorry I mean rent it out) again.
2. For you to be negatively affected by this, the advertiser doesn't need to know who you are. If you try to buy an item and they hike up the price for you because they have your browsing history and know that you are very interested in that kind of item, it doesn't matter if they know it is "samcraig" or not.