If you were interested in seeing any details whatsoever about the protocol there are none either in the article or on the official website.
https://silentcircle.com/web/scimp-protocol/
I thought we didn't want to use the NIST curves (like ECC-384?) anymore... seems something like Curve25519 might be better.
http://silentcircle.wordpress.com/2013/10/17/this-one-goes-t...