The difference I can see is that a filesystem operates down at the kernel level, and Microsoft would have to prove that they had not turned Windows into a derivation of btrfs. But if Windows booted from FAT or NTFS and then loaded a btrfs module from its boot partition, without any btrfs code or derived code being compiled into Windows, then why not? (Obviously, they'd have to comply with the GPL and offer the source of their btrfs module).
Whether that's do-able, I have no idea. Windows source is secret, btrfs source I haven't read. Related - there's the Linux network module which encapsulates and executes an NDIS driver written for Windows under Linux. A kluge, but no legal challenges that I've heard of.
In passing, I'd love to see a port of Linux LVM to Windows - probably less of a technical challenge, though I'd guess only MS could do it and MS probably has many reasons not to.
Copyright © 2017, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds