This and the prerequisite doc, draft-ietf-alto-unified-props-new, seem to adequately call out differences v. base ALTO (rfc7285).  

I'm somewhat concerned about the building of what's looking like an overlay routing system protected only by pairwise relationships between the parties rather than by authenticating the data (as BGPsec aims to do), but that's a much broader architectural concern, which does not directly impact this extension.