The problem was Windows giving arbitrary access to the kernel to software that can be updated OTA without user intervention and allowing that to crash the kernel, right? Wouldn't this mean that Windows is considerably less secure and stable than assumed?
Not really. These were kernel modules authorized and installed by the system admin. Of course kernel code runs the risk of crashing your system. The same is true on Linux, and according to another commenter it already has happened with Crowdstrike for Linux
The problem was Windows giving arbitrary access to the kernel to software that can be updated OTA without user intervention and allowing that to crash the kernel, right? Wouldn't this mean that Windows is considerably less secure and stable than assumed?