Yes! The only real solutions are client-side: IP-block trackers and encrypt your connection. The "please, sir, can I have some privacy" nonsense of do-not-track headers and IAB opt-out pages needs to die.
Even those actions have limited value as long as you have any other reasonably reliable signature as you navigate the Web. Anyone with a static IPv4 address for their home computers does. Anyone with a browser that will allow queries for a bunch of collectively-almost-unique properties of the host system does. Even certain standard web protocols inherently act in this way when used for their intended purpose.
That means almost the entire web-using population is carrying around at least triple signature information, and if too many people start to block ads and trackers served via third-party systems, the major networks providing those things will just move to a model where sites hosting their ads act as proxies and serve the content from their own domain, which if anything would be slightly worse for privacy.
This risk will remain until we get fixes for each issue I mentioned above. Leaving aside TOR and the like, we could move towards dynamic and rapidly rotating IPv6 addresses as points of origin with sufficient ISP support. Browsers could then close the other loopholes, but some of them will be difficult to fully eliminate and keep eliminated without compromising the user experience, because unfortunately some features that are useful for legitimate purposes are also inherently leaky.
Making spying expensive and difficult is expensive and difficult, because most people are transparent. Think of how much work it takes for an espionage service to maintain an agent in a foreign territory. It's not good enough to have dead drops and cut-outs and safe houses; you need an entire network of them so you don't create patterns, and then you need to compartmentalize it to guard against the inevitable compromises.
The problem for privacy advocates is that encrypting everything itself sends a signal, albeit an obfuscated one. The solution ot this is 'everyone encrypt everything' but most people simply do not have the motivation to do so. Notwithstanding the 'everyone has something to hide' argument - most people don't. Either their private matters are unimportant, or they're not important enough to be leaned on for anything.
Further, from the standpoint of a bad actor (eg an overbearing state or any private extortionist), you don't need to penetrate people's privacy to compromise them, eg 'Help us out even though you are unwilling to do so, or we'll spread the word that you're a pedophile' would be enough to suborn many people's cooperation, without the accusation having any factual grounding whatsoever. In western society being labeled as a pedophile is so bad that most people turn off their brain as soon as they hear the word and are willing to impute guilt automatically. In fact, they don't even need to get the word right: http://www.theguardian.com/uk/2000/aug/30/childprotection.so...
This means you either have to eliminate the hostile party by buying or replacing them, or you go around them by providing en encrypted tunnel. And doing that reliably means doing away with MITM that browsers trust.
And is privacy only for people who can afford to spend hundreds per month on it?
The carriers might get a brief sugar rush on their EIBTDA initially (from the monetized, non-private traffic), but over the long hall this is going to be a negative growth market. The consequence is it hits google at center of mass.
You can't set a precedent of paying for privacy because you can't pay everyone who can spy on you. All it takes is for one reasonably frequently used service (credit card, Amazon, AdWords) to refuse to respect privacy, and you are back to square one.
Privacy is not something a single company can successfully sell you by paying other companies not to spy. It can only be obtained by making spying expensive and difficult. (Perhaps with cryptography, perhaps with legislation, perhaps through some other means)