It is entirely feasible to build a web browser which ensures privacy, but it would come at a steep cost. If you disable the "REFERER" [SIC] sent by the browser, a lot of sites break. If you disable the "USER_AGENT" sent by the browser, a lot of sites break. If you disable "COOKIES" in the browser, a lot of sites break. If you disable "FLASH" in the browser, a lot of sites break. If you disable "JAVASCRIPT" in the browser, a lot of sites break. If you require PROXIES or Onion Routing, a lot of sites break.
With mobile devices, disabling GPS/Signal location, network access, address book, and many other features will break a lot of sites/apps/services.
I'm sure you can see the pattern.
The tough question is, would you use a browser/system that intentionally broke most of the web/Internet to ensure your privacy?
Do you want to be the person fielding support calls from your mom when some company on the `net insists on having privacy invading misfeatures enabled?
The most "ambitious startup" might be in trying to fix the privacy and security problems of the existing Internet by refusing to support the regularly misused misfeatures, but the resulting loss of functionality and compatibility might be too much to bear.
> We should be careful not to undercut the economic underpinning of the Internet in our attempts to regulate online privacy.
Whether regulated by new laws or prevented by technical means, you would not eradicate the economic underpinnings, instead, the underpinnings would simply change. For example, people would still pay Google to display advertising along with search results even if it was impossible to do user-identifiable profiling. The untargeted advertising might be less effective than targeted advertising, but it would still be far more effective than no advertising at all.