Structural functions may be better for what you want because the thing that it does well (better than other books I have read certainly) is explain the sorts of harmonic substitutions that start happening as you get into the late romantic period. He believed (“Theory of Harmony” is also like this) in explaining harmonic theory entirely from the perspective of composition so it explains from first principles what you see composers actually doing rather than trying to find rules and then examples of those rules. So in “Theory of Harmony” he builds up an entire theory of harmony starting from simple 4 part voicings and voice leading, in “Structural Functions” he sort of condenses that into a couple of very quick chapters and then moves on to talk about key relationships in more depth.

It also breaks sharply with the old orthodoxy where authors used to talk about everything as key modulations and then struggle to explain things which were really related to the original key area but seem very remote in the new “modulated” key. What Schoenberg points out is that music actually very seldom actually changes key. In fact almost the only time this happens is in the classic “Whitney Houston” where a pop tune just goes up a tone for the last chorus and never goes back. Music in the classical tradition almost always ends up in the same key that it starts in even if it visits some remote areas along the way and therefore as Schoenberg argues it makes sense to understand key areas that are temporary in relation to the original tonality. It also means you end up with a harmonic theory that works perfectly to explain jazz standards and pop music.