This question contains a slight misunderstanding of the UCMJ. Article 117 deals with “Provoking speeches or gestures,” not threats. A threat is typically prosecuted under the General Article, Article 134, UCMJ, which prohibits all disorders and neglects to the prejudice of good order and discipline. Communicating a threat is a classic example of such conduct. The use of an encrypted messaging platform presents an evidentiary challenge for the government, but the offense is still prosecutable.
To prosecute the offense, the government must first gain access to the threatening messages. This is the biggest hurdle. Investigators, such as CID or NCIS, may be able to obtain the messages through a search warrant for the electronic device, by obtaining consent from the recipient of the threat, or through other investigative techniques. Once the content of the message is obtained, the use of an encrypted platform is irrelevant to whether the words constitute a threat.
The prosecution must then prove beyond a reasonable doubt that the accused was the one who sent the message and that the language used constituted a “true threat.” This means the message, when viewed in context, would convey to a reasonable person a serious expression of an intent to inflict bodily harm. A military defense attorney would challenge the attribution of the message to their client and would also argue that the language used was a joke, hyperbole, or was not a serious expression of intent to harm.