Proving how difficult it can be to talk about security features in plain mainstream terms and remain technically correct, the article itself (while explaining how Schmidt was wrong) claims something that ranges from seriously misleading to just plain wrong depending upon how literally you read it:
"Incognito mode deletes Chrome's browsing history and other detritus that help track users around the Web. Chrome users who deploy incognito mode can, for example, prevent other people who access their computer from seeing their search history. Importantly, this deletion of local history logs does not at all affect external entities' access to said history."
Incognito mode doesn't delete Chrome's browsing history, it suspends it from keeping new history for that tab or window's session, which is quite a different thing.
Then again, the cynic in me would argue that Mr Schmidt knows perfectly well the limitations of incognito mode.
If Firefox drops a tormode soon, they are going to blow Chrome out of the water.