it looks to me as if there are 2 key pull requests that all the remaining ones depend on. although one of those is in the secp256k library, and it constitutes the hot new-ish cryptography that provides some of the cool properties that will make these encrypted connections so difficult to fingerprint. i'd expect anything like that might take many months to get merged (it's ~6 months old as of now).

however, I'm pretty confident it's viable, the secp256k devs probably aren't wasting any time on that kind of work if it's not a serious spec/upgrade