It's tough to argue with success. They obviously wanted someone who was going to go all-in on the cloud angle when they picked Nadella, and to give credit where it's due, the stock price chart for Microsoft says everything you need to say about the results he's delivered.
What's worrying in a broader sense is that our societies have come to depend heavily on Microsoft's previous flagship offering, Windows. As the article observes, the focus has shifted elsewhere under Nadella's term. And as many people can attest, Windows 10 is far from the Windows we used to rely on, particularly for those not in enterprise world or academia. Many millions of people are now fundamentally depending on a product that is no longer the goose laying golden eggs for its maker, and the maker's economic incentives are no longer aligned with the best interests of their users.
Windows still directly generates annualized revenue of $22 billion. Plus Microsoft's dominance in that platform creates a ton of synergies in its other product lines. For example how many Azure customers pick it over AWS because they need a Windows-first cloud. It easily contributes hundreds of billions to Microsoft's market cap.
I'm sure Windows is no longer considered the sexy department in Micrsoft's internal culture. But corporate management would have to be ridiculously incompetent to let that business line get significantly neglected.
This seems like a classic case of enterprise vs. small business vs. personal use. The serious money is at the enterprise end, and most of the effects you described will be most profitable there. And of course the Windows 10 editions for large organisations don't try to play the same games with things like forced updates and mandatory telemetry, because they know very well that no professional IT department supporting thousands of users is ever going to surrender control of its equipment like that.
Still, the reason the makers of expensive professional software have largely turned a blind eye to privacy forever and/or given away very cheap or free copies to students or home users is because they're well aware that it's a huge advantage for their software to be the cultural default. For example, would Adobe really have become dominant in digital content creation software if loads of enthusiastic design students hadn't been honing their skills on ripped off copies of Photoshop? Would Microsoft have come to dominate business desktops if home and school PCs hadn't all come with Windows so everyone already knew how to use it, or if they hadn't given away their development tools mostly for free to anyone who wanted them? We'll never know for sure, but it would be naive to discount the effects of familiarity.
This feels a bit like Apple's recent trend for messing up its laptops and largely ignoring the rest of desktop/server space other than producing an occasional statement piece. Sure, they make the vast majority of their money from their mobile devices and the surrounding ecosystem, but someone still has to write the apps that drive that ecosystem, and those people are primarily using Apple laptops. If you ruin your development platform, and the developers start shifting their efforts towards other platforms (presumably Android in this case) then your users will follow, and now your business model is dead, even though the part you neglected was only directly responsible for a tiny part of your revenue stream.
Obviously the danger isn't quite the same for Microsoft because strategically it's banking on cloud rather than mobile for its big revenues. But even so, cloud hosting is a pretty competitive space right now, and the Windows-friendly integrations (for developers as well as users) are a competitive advantage, one that is at risk if the very long tail of smaller developers start getting fed up of Windows and shifting to other systems. I suspect Microsoft's leadership are well aware of this danger and the often better developer experience on Linux is a big driver for introducing WSL, but still, if that's all they've got then you can just as well use a real Linux box and not pay the Microsoft tax or deal with all the unwanted games that Windows 10 keeps trying to play. If Windows 10 becomes too much of a liability for small businesses and power users, sooner or later that will inevitably have consequences for the whole business model and all those synergies you mentioned, even if it only happens over a period of years.
As long as important software still runs on Windows and isn't available on other platforms, the answer will clearly be yes. In time, particularly with the shift towards online systems where the browser is now playing the standardised platform role that desktop operating systems used to play, hopefully the dependencies on Windows are becoming fewer and weaker. However, I think we're still a few years away from being able to assume that anyone can move off Windows without losing anything important to them.
What's worrying in a broader sense is that our societies have come to depend heavily on Microsoft's previous flagship offering, Windows. As the article observes, the focus has shifted elsewhere under Nadella's term. And as many people can attest, Windows 10 is far from the Windows we used to rely on, particularly for those not in enterprise world or academia. Many millions of people are now fundamentally depending on a product that is no longer the goose laying golden eggs for its maker, and the maker's economic incentives are no longer aligned with the best interests of their users.