Facebook will try to track me and my behavior, and tailor ads based on it, regardless of whether I'm a Facebook user and regardless of whether I'm logged in. In theory any link someone sends me might end up trying to connect to Facebook's advertising network and record information about my behavior.
At what point do I "agree" to this? What do I get "in return" for this? What if it's a crappy article that I close right away? Or the page load takes too long and I close it before it finishes (but, in true form, the ad code would already have run)? Since I don't know until I've loaded the page -- at which point you think I have an ethical duty to let Facebook serve ads -- whether it will have ads at all, or whether they'll be Facebook ads, at what point do I find that out and have an opportunity, in your model, to decide I don't want to be a party to that?
(in case you're wondering, the correct answer is "use an ad-blocker to pre-emptively ensure nobody can force ads on you without your agreement")
I'm talking about being a Facebook user, as in maintaining a profile and logging on and using the site. My contention is that by affirmatively choosing to use their resources, an ethical obligation is created.
There is no coherent position which creates an ethical obligation to view sites in full and unmodified (especially given that one of the primary purposes of web browsers, historically, has been to mangle what a site sent to you into something that's actually a sensible and renderable document tree).
Practically: because it leads to a race to the bottom, lowering to their level, the world just gets worse and worse
Philosophically: because ethics are a fundamental part of who you are, and strangers acting unethically does not relieve you of their burdens