The risk of large scale surveillance is that it can generate data sets that can be mined for information. Tracking can show networks of friends, attendance at political rallies, books read, movies watched, foods and alcohol consumed. Does this pattern match for a potential terrorist - can't prove anything, but maybe you shouldn't keep your job at Lockheed, or should get extra screening at the airport? Did you watch "little miss sunshine" too many times for your demographic - could mean you are a pedophile - maybe you shouldn't have a job as a school teacher - think of the children.
Which political information should you see? Candidates can target their adds to YOU specifically. Same for news, and advertising.
Maybe you don't get enough sleep, or are found to meet women ( or other men) at bars and take them home. Sounds like "statistically" you might be a health risk and your insurance rates will go up.
Large scale tracking, data collection and analysis allow for statistical pattern matches. The public might be happy that a new system has a only 1% failure rate, and only a 10% false positive rate for recognizing people who are a danger to children - unless you are in that 10%