It's not a bad PR move, and it will probably ease some advertiser's concerns. However, there will be other advertisers who will be annoyed to see that the number of invalid clicks Google found doesn't match either the number of invalid clicks logged at their own sites and/or through a third-party tool (if they're using one).
For the times at AV when there were click count discrepancies, I had a mantra: "A click is not a clickthrough." There are numerous reasons for discrepancies; too many for me to type in here. In fact, even before click fraud became a serious problem, I was against the use of "redirection" as a method of establishing "who went where" after they left our site, because it offered no guarantees that the surfer actually got there, and just increased our processing requirements.
An irony of click fraud detection is that the better it gets, the more money you lose. You find out what you can't charge for, and you spend money in the process. It's a lose-lose situation because the fraudsters keep coming back. (If there was a better way to determine the "identity" of a clickstream, perhaps improved click fraud detection would discourage fraudsters, because it would increase the likelihood that they'd be caught. However, pretty much everything sent to a web server can be faked.) All you do is just play cat-and-mouse games with the fraudsters. It's another reason why I think CPC is a bad business model.
There have been arguments made that increased click fraud detection will actually save money for advertisers. That's not so clear to me. It will be necessary to maintain and improve existing click fraud detection to give advertisers assurance (or at least hope) that click fraud is being detected. That will cost money that has to come from somewhere. If it comes out of profits, the search engines and ad networks will lose value, so they will have to pass the cost of improved click fraud detection on to the advertisers.