I think Google still innovate in that, as other commenters have noted, they're creating some interesting file system and database solutions. However, these (among others) are infrastructural contributions to the backend. I think the article is talking about innovation from the average consumer's perspective.
I remember speaking to a friend of mine when Google released Wave. He said Wave was the first real home-grown "product" since GMail and it was his opinion that it was developed in Australia because the office there was a kind of skunkworks plant with little bureaucratic oversight compared to the other Google offices around the world (whether this is true, I don't know). He said Google were at the stage where they are big enough for middle management to stifle real innovation because the individual managers don't need to take the kind of risk required with greenlighting a project like Wave (and look how that turned out!).
To Google, buying an early-stage startup is a cost-effective way of outsourcing the risk-taking and that kind of acquisition is the closest they can now get to serious innovation and staying competitive.
From the average consumer's perspective, innovation doesn't mean squat. Consumers don't care who invented what, they care about the products and services that are available to them.
The Go language, Chrome OS, the recently announced redesign of the backend storage architecture that's supposed to replace MapReduce for some tasks. These are the things that I remember off the top of my head without going through the stuff they keep announcing all the time.
[Edit:] Things like native client and V8 might also be worth mentioning.
Apart from these, Google has come up with the WebP image format which may make internet significantly faster than what it is today.
Furthermore Google hosts jquery files on its servers, which for many people is a godsend. While it is a simple thing to implement, it ranks highly on how it benefits websites (and thus the end users of those websites)
Google didn't invent page rank, they stole the whole thing in 1996 from a research project by Larry Page and Sergey Brin, two kids at stanford. Have they even seen a dime of money since then?
So John Doerr, a venture capitalist, "lost" an impromtu argument at a Web 2.0 Summit and this is supposed to mean that Google and Facebook don't innovate?
And also, all the technologies created for distributed computing by google i.e. Mapreduce, GFS, BigTable. Currently most of the web is using Hadoop, HBase, Cassandra and other tools which are all inspired by google work.
I would prefer to call it an operating environment. Not that different from many others - .Net, the DOM environment of JavaScript, any scripting language in any larger tool.
Lets use "operating system" to mean what its generally understood to mean: something that you boot on hardware, that manages system resources/hosts drivers for devices, manages time and memory.
Just a separation-of-process model is not enough to justify the name Operating System.
Wasn't Google planning (or had in beta?) a voice service before buying Grandcentral? It just seems that when you get to the size of Google, if you see a market that you want to get into and that has an easy to acquire, quality product currently being created, why not buy it? You get the double bonus of a working product and some talent.
Facebook Platform, Graph API, Like Button, Social Applications - a few truly transformative things that have come out of facebook in the past few years. Saying facebook doesn't innovate is rubbish. Though they do tend to borrow an idea here and there, they have sometimes left it better than they found it (exceptions of Quora and 4sq)
I remember speaking to a friend of mine when Google released Wave. He said Wave was the first real home-grown "product" since GMail and it was his opinion that it was developed in Australia because the office there was a kind of skunkworks plant with little bureaucratic oversight compared to the other Google offices around the world (whether this is true, I don't know). He said Google were at the stage where they are big enough for middle management to stifle real innovation because the individual managers don't need to take the kind of risk required with greenlighting a project like Wave (and look how that turned out!).
To Google, buying an early-stage startup is a cost-effective way of outsourcing the risk-taking and that kind of acquisition is the closest they can now get to serious innovation and staying competitive.