I basically agree that what Amazon (and Google for that matter; disclaimer: I work for Google on display ads) is intent and intent is huge. Much as Facebook would like to become the de facto Internet, I don't see it ever capturing that intent (let alone Twitter).
That being said, I think people discount what an inordinate amount of good advertising has done and just focus on the negatives (eg the "shit you don't need"). For example:
1. While you might be able to afford a few dollars a month to pay for the cost of GMail, as one example (note: I have no idea what the cost would be here; I'm just making up numbers), would you? People are highly reluctant to spend on this. But more to the point, what about people in Africa or Asia for whom a few dollars a month is a significant sum of money? Those people can enjoy the same service with an ad-supported model; and
2. Traditional advertising channels (print, radio, TV) weren't affordable to smaller businesses. Things like Adwords have allowed many businesses to exist that otherwise couldn't before the advent of online advertising. How many jobs do you think this has created and supports?
But I do see the whole "eyeballs-then-advertising" SV model as peaking if it hasn't already. Facebook may well be the 800 pound gorilla that broke the camel's back on that one.
And while the author notes that advertising space is growing faster than purchasing power not all advertising space is created equal. What you simply has is a diversity of distribution channels for your advertising. That's a good thing. But not everyone will succeed with an ad-supported model, that's true.
To the author's problem of working on big problems, I agree. The fact that SpaceX spent less developing a launch system and reentry vehicle than was spent buying Instagram [1] with its 13 employees makes me sad that so many bright minds are working on the apocryphal social network for cats.
[1]: I realize with the 50%+ drop in Facebook's IPO price, that sum is now significantly less.