Most people do not need to use typical office software at home. Those that do are often supplied a work laptop so the company has control of the hardware and installed applications. Other than a few specialist applications and games most people do not need Winbloat or any other specific OS at home.
What most people need in an OS is to support a main stream browser, say Chrome or Firefox, is relatively easy to update, has photo/video/audio viewing/listening capabilities, and similar activities. These activities can be done on a smartphone, tablet, laptop, or desktop with currently available OSes including iOS, Android, ChromeOS, Windows, Linux, and BSD - I am probably omitting a few. There is little value to most people in being in any ecosystem. In fact being very cloudy makes the OS less critical. Linux users can use Office365 or Adobe Photoshop if they are willing to fork over a subscription.
Thus the existential threat to Winbloat, beside Redmond stupidity, is most users do not specifically, objectively need it. They need an OS and Winbloat often comes preinstalled. For many this is acceptable - they have a functional device out of the box. But if the OEM's decide to push an alternative OS, say AndroidOS or ChromeOS, most users could migrate away and many will if the price is right. And some will.
Many longish term navel gazing is that Winbloat marketshare on mobile will be low (~5%) and its share will drop on the desktop market to about ~70% with commercial Linux (AndroidOS, ChromeOS) being the main beneficiaries. Also, the PC shipments will decline for a few more years then level off at natural replacement rate.