It's also been available under Apache 2.0 this whole time, which means you get a patent grant that's less problematic than the one Microsoft is using for .NET Core. So anyone looking to work on a .NET-ish runtime can do so without fear of litigation by starting with the Micro Framework sources. The caveat is that the license only applies to the things that are already in NETMF. The moment you try to start adding things (e.g., by ganking sources from CoreCLR), you're back in the murky territory of coming up with an answer to the question of whether or not the MIT License is enough to keep you from being on the receiving end of patent litigation.
The irony is that fragmentation is exactly what Microsoft is trying to stave off, but the policy they've chosen with their patent promise does its part to sort of force the hand of interested parties in that direction.
Is there a meaningful ? I.e., what does classical inheritance get you that prototypes-under-the-hood don't? For a programmer using classes, that ES's are implemented using prototypes should be as important has how the JVM works.
Microsoft will need to revise their patent pledge first. If you look at the way it's currently written, it's almost as if it was specifically meant to preserve their option to pull an Oracle v Google against a vendor who decides to pick up .NET and run with it.
Meanwhile, Google has switched away from their Apache Harmony derivative to the officially anointed sources, so there's not much incentive to make the move away from Java entirely, anyway.