Say you are a bread maker, do you care the shopping environment your customers buy the bread, if you only pay commission to the shopping mall according to the sales number? Of course you don't, because if one shopping mall doesn't sell any of your bread, you have no cost!
So yes, ad publishers may not care about Google users in and of themselves, but they care about volume, and Google making its users happy and engaged keeps that volume high. By transitivity, ad publishers should (theoretically or indirectly) care about whether Google treats its users well as it would have some positive correlation with their viewership volume.
You're trying to draw a false dichotomy between the two -- they are not mutually exclusive goals. Just because Google is building for the users doesn't mean it's not building for the ad publishers. By building and improving their products for the users, they are increasing the stickiness and value of their users for the ad publishers.
If you had to choose between advertising at a location where people just breeze right through, and advertising at a location where people stick around to browse the goods, which location would you rather pick? Obviously the latter where people stick around, as it's more likely that they'll notice your advertisement. Google improving their products for the users makes it more likely that their users will stick around to notice the ads.
Assume it's true, then Google 1) will definitely abuse the position, and 2) for Google, the top priority is about how to maintain the unique position, which has less to do with users' experience/privacy/(put anything here), but more to do with the browser + search engine paradigm. In fact, with the emerging of Smartphone + Walled gardened Apps paradigm, it's about the time to predict the decline of Adwords.
Say it's false... well, so is your argument.
On the other hand, your argument that Schmidt's post is "b.s." hinges on the assumption that catering to users is a completely separate business model distinct from catering to their direct revenue sources (e.g. ad publishers). It's not.
The fact is, Google occupies a unique position under the framework of browser + search engine. Their inability to expand meaningfully to other web segments (except gmail) proves:
1) they are idiots. (highly unlikely) OR
2) the theory of "user engagement" does NOT work in case of Google Inc. Therefore, mining of users' data serves purely for the business interest, it has no value added to users. And yes, Google Search's model is fundamentally different than ugc-centric services (e.g. fb).