More faster than saying “Look mom: No mac address needed!”
If during a browser-specific session a user queries a search engine and accesses subscription-based web services provided by the same search engine (web-based email accounts, gadgets, apps…), the IP used for searching can be associated to his web service credentials (username, password…). Therefore, it is possible for a search engine to guess the identity of that user and know what the user is searching for, when, and how. With referrer and click-through data, it is also possible for that search engine to know where said user came from and where did he/she go.
In most cases, geolocation data are much more accurate for devices with GPS, like smart phones, and HTML5-compatible browsers. In general, users privacy becomes increasingly compromised as more web services, apps, and device features are enabled. On the Web, most free stuffs are not really free, but involve a privacy cost; otherwise, they won’t be free.
Of course, if during a session a user lends the device to another searcher, the search engine might not be able to guess the identity of that user.