An obvious way this can harm Linux is by giving junior users coming from a windows environment a way to do much of the things GNU/Linux can do - even if it's potentially inferior (would they even know?) - without actually becoming a user of anything running the Linux kernel.
Do you expect the Linux kernel to continue being runnable on contemporary laptops if decreasing numbers of users are even attempting to boot it?
The same thing happens with Docker on Windows in the datacenter. If users embrace the MS offering, and it runs their Linux containers, there will be less resources going into Linux kernel development from the server world.
Google going the fuscia/zircon or whatever it's called now route further diminishes the resources put into the Linux kernel.
I don't think there's any reason to reject the observation that there's a lot of competition in this space and a number of these moves substantially threaten Linux's relevance in the long-term.
The exodus of users to OSX has already been quite harmful to GNU/Linux's progress on the desktop. If it weren't for Intel investing so heavily in Linux kernel development I don't think we'd have modern wifi or gpu support, not from the community alone.
Much of what I describe above is the Embrace phase. Once a sufficient number of users are running their Linux userspace components on proprietary underlying kernels, the relevant companies can start investing in "improving" userspace in incompatible ways only their kernels implement - perhaps in patent-protected ways - Extend. Voila, you're locked in. And since the Linux kernel wouldn't have the development resources it once had, it just falls behind into obsolescence.