Much of what privacy advocates focus on in terms of privacy violations on the internet neglect to take into account privacy gains from those same technologies. A key factor to this is focusing on what information is being shared but neglecting to account who it is being shared with and the privacy benefits associated with interacting with that service instead of traditional alternatives. For example, Google can see every search a person conducts regarding their sexual preferences. However, a closeted homosexual might turn to Google to find social and support groups because they're afraid of judgement from their family and peers. Groups like the EFF would say that buying a book through a Kindle is awful for the privacy of the reader because Amazon tracks the fact that you specifically bought the book and can see when you're reading it and even what page you're on - but most readers don't care that Amazon knows they're reading 50 Shades of Grey; they'd rather that everyone else on the subway just sees them reading their Kindle instead of an erotic novel paperback.
In general, people are more afraid of having their sensitive personal details disclosed to the people they interact with in person - their families, friends, coworkers or even the checkout lady at the pharmacy. The loss of privacy to a remote party like Google is worth the net gain in privacy to avoid disclosing embarrassing information to people they are more concerned with. This is also dependent on the type information being disclosed - the set of information that you'd want to keep from your friends is not the same set that you'd want to keep from your spouse, which is not the same as what you'd keep from your coworkers, which is not the same as what you'd hide from the police.
For those that would prefer an audio version, Ben Wittes gives a good discussion of it on the Lawfare Podcast at http://traffic.libsyn.com/lawfare/Episode_129.mp3