Disclaimer: I'm not an apple employee and not am not lawyer, this is just my understanding of the situation having spent lots of time talking to lawyers and other experts about this:
I agree with much of the sentiment that software patents are often silly and the system is broken in many ways. This patent is a reasonable example of that (patenting syntactic sugar for monads, really?). I have no idea if there is prior art, but I wouldn't be surprised. For sake of discussion, lets assume the patent is valid.
Even if I and others don't like it, the software patent system exists. As is pointed out upthread, one of the major reasons that Swift uses the Apache 2 license is to provide more certainty for the community w.r.t. licensing and patents. An additional bonus of the Apache 2 license is that the open source project as a whole benefits from companies having and contributing their patents under the terms of the license: to say more directly, it is good for the Swift project that Apple has this patent and has contributed it to the project.
The reason for this is the Apache patent revocation clause, the end of bullet 3:
If You institute patent litigation against any entity (including a cross-claim or counterclaim
in a lawsuit) alleging that the Work or a Contribution incorporated within the Work constitutes
direct or contributory patent infringement, then any patent licenses granted to You under this
License for that Work shall terminate as of the date such litigation is filed.
This basically says that if someone sues someone else over Swift then they lose access to the patents contributed to the project, and are therefore subject to countersuits. This is a significant part of the protection that the Apache license provides (it is a big deterent to lawsuits in general) but it only has teeth if there are actually patents in play!
The LLVM.org community is actively working on a multi-year relicensing effort specifically to achieve these sorts of protections for LLVM as well.