Google was "just" another search engine, sure. But it was far better than any of the other search engines out there. You can't say that of the groupon clones. They're all slightly different ways of doing Groupon, and none of them at this point look like they're offering any significant advantage over eachother. They may be more or less sustainable or profitable than Groupon, but they aren't game changers in the sense that Google was.
She wasn't complaining that startups have exits, she was complaining that startups are being designed around exits. None of the counter-examples you listed, nor any established company that anyone can name off hand, was designed for an exit. You can look at the guys who did <s>gmail</s> Google Maps* or Youtube and say yes, they got a good exit, but both companies were designed to do something new, and their exit followed from that. A lot of startups these days seem to be designed to show off design skills and the ability to write an app, rather than actually do something unique and useful.
Which brings me to your last point. Writing Basic interpreters for hobbyists wasn't trivial in the sense that the author meant. Finding a cupcake or sharing a photo is trivial: it doesn't much change anyone's life, and it's been done a thousand times before. It's trivial in the sense that a fart app for the iPhone is trivial. Microsoft's innovation was not that it could write basic interpreters, as people wrote basic interpreters for new architectures all the time. Their innovation was that they were an independent software firm who contracted their services out to hardware manufacturers. This is how they came to land the DOS contract with IBM that catapulted them to success. That innovation was as big as the innovation where someone first said, "why are we hosting this on a server in the basement? Why don't we throw it up on some server out there that doesn't care where we are or what device we're using to access it?"
If you look at any of the big tech companies around today, they got their start by doing something either a) no one had done before, or b) an order of magnitude better than anyone had ever done it. Most of the startups you see today don't make any effort at a) or b), but they are c) designed to be flipped in a talent acquisition. This is what the author was complaining about, and if you don't accept it as a legitimate criticism then you at least have to address her actual point.
*Edit: see response from protagonist_h