On reading the headline, I was extremely skeptical of the article (Google is like 90s Apple? Surely not). I am still not convinced, but DED makes some really good points.
The massive issue for Google is that its entire operations are bankrolled by web ads. These can be vulnerable to several fronts:
1) Monopoly accusations like those faced by MS.
2) Increased regulations preventing them from capturing as much user data.
3) Customers shifting to no -Google search due to privacy concerns (DuckDuckGo) or because browsers replacing Google search with alternates.
4) A new revolutionary competitor who slowly erodes Google's lead
5) Apple and MS eliminating the need for Google searches through better and more prevalent apps.
While I don't think any of these are likely, the point is that Google as a business has a Death Star like single point of failure. One which may be attacked by the natural and ongoing desktop to mobile transition even if none of the revolutionary scenarios above come into play.
That being said, Larry Page is a smart guy and I am sure aware of this issue. Google must be actively involved in fighting this off.
2) You only get regulated when you do unhealthy things. Invading privacy (Google probably knows as much about you as the NSA does, even if you never had a gmail/G+ account) is one way to get regulations imposed on your business.
3) Honestly I don't see this happening at all. Google will remain the king of search as long as nobody does it better, and nobody can really improve upon Google without outright making it work like Google as a starting point to begin with because that's how all the SEO spammers program their bots. Remember Lycos? Altavista? Those originally started by indexing keywords found in pages, but using those engines back before google was the worst user experience ever. Yahoo! tried to do it better by putting the human factor back into it, but they are obsessed with making the user experience more like a newspaper (eg horrible) than actually providing anything usable, so they also went away. Bing and Yandex use Google-like algorithms as a starting base, so they start out with google-like results, but ultimately they don't differentiate themselves from Google at all, and that's why they aren't used as the first option. It's the non-search things that make people stick around. I find Bing's translation is a bit less lol-worthy than Google, but I end up using both if I need to stumble through a foreign website.
4) Not happening until the next "thing" comes along People kinda haven't noticed how break-neck of speed we moved from electricity to radio, to television, to computers, to internet, to cellular. Mobile internet is like the final form of all these technologies and Nikola Tesla would love it. All in less than 150 years. When you look at the software side, it's not so dramatic, as most software will be unusable before it ever enters the public domain. Will Google still be around in 20 years? It wouldn't be if it was just a search engine. It has to remain relevant.
5) Microsoft I'm confident can keep a #1 or #2 place if it can push Google out of that position for the search engine, email, or even mobile devices. The problem is that Microsoft isn't Apple and isn't producing it's own hardware that runs it's own OS. The Surface Pro is the only device that does, and the only Surface Pro device that anyone would want is the Surface Pro 2. The 3 is just a cheapened version of the SP2, and the RT doesn't run x86/x86-64 software, thus confusing everything. What Microsoft should have done was made the RT their "iPad", and kept the classic desktop for all other x86 versions, keeping the "touch mode" for convertible laptops. That would have sold it. Too late now, but maybe they should have thought of this when they came out with Windows Phone/Mobile. Android has so far been a total failure for anything but phones, and that's because the OEM's are using it as a throw-away OS to power weak Android phones and trying to pass them off as being as good as an iPhone.
Honestly if you look at history, either Google has to step up and make their own hardware and OS like Apple does, or Microsoft has to step up and make their own hardware. Here I'll even offer an obvious compromise everyone could be happy with:
Apple, Microsoft and Google could all offer their OS for free (closed or open, doesn't matter,) and provide just the documentation needed to write drivers for the OS, but the hardware manufacturer has to support the hardware for 10 years by providing the drivers for all versions of the operating system that is released after the device is sold, and can not sell hardware of inferior performance to what the first party sells themselves. The idea is for OEM's to step up and produce hardware that people want, not "cheap clones" which provide poor experiences.
There are plenty of people out there who don't buy a Mac because they want a desktop that they can put a SLI video card configuration in, and they'd dump Windows in a heartbeat if there was an option to. Steam on Linux doesn't fill this gap because making a Linux desktop work is as horrible as pre-plug-and-play was in 1993. Google could come out with a desktop OS, (ChromeOS and Firefox OS aren't OS's, they're Web Thin-clients,) but they would have no incentive to since nobody is demanding it.
I don't see Apple ever licencing their OS again, even though there are at least two markets it would be perfectly suitable to that wouldn't result in a repeat of the Mac Clone era (Rack Servers and High-end gaming systems) that Apple completely ignores. There are entire communities dedicated to doing this without Apple's blessing. If the pirates put as much effort into hacking the OS they could probably write a new OS that is compatible with OS X and that would be a serious problem for Apple.