Actually, that antenna efficiency (percentage of power that is actually radiated) is a useful number and we use it in many projects. It's easy to obtain from simulation, not so easy from measurement because you need to measure the pattern over all directions.
So I don't agree with vfone: it's nothing theoretical, it's a practical and useful figure of merit for antenna performance.
What you call "terminal efficiency" is trivial: S-parameters are voltage ratio, so normalized reflected power is |S11|2 and your requested efficiency is 1-|S11|2
But as mentioned, that has nothing to do with antenna efficiency, which is defined differently.