Great question! When we released Swift for TensorFlow to open source last year, our primary focus was (and still is) to work with the machine learning community because we believe the ML community is critical not only to TensorFlow's success but also to the success of Swift in the machine learning field. Having open discussions of Swift in the TensorFlow community helps grow interest in Swift among ML researchers and students. Therefore, we chose TensorFlow's forum (firstname.lastname@example.org) as the primary site for discussion, and encouraged people to discuss everything there. I've seen many people from the Swift community join discussions on that forum, and I think it's great because it is bringing new perspectives to the ML community.
As the project grows, I think it is possible for Swift for TensorFlow to be in "Related Projects". It ultimately depends on projects leaders on both Swift.org and TensorFlow sides to decide what's best for the project and for the community. @tkremenek @Chris_Lattner3