Why are you saying that the chords in brackets are in the key of F? At that stage, the progression seems as though it's going back to C after a brief detour to Dm (ii in C); so the large-scale progression, up to the G7, would be I–ii–V7 in C major, with the ii elaborated by being tonicized. From about the A7 onwards, however, the progression could fit into an F major context (as opposed to, say, G major), which is why Schoenberg says that 'F major is in effect' at that point and so it's from there on that the music moves towards F, which is then confirmed with a cadence.
G7–F/A isn't a cadence: it's a deceptive cadence, which means that it sounds as if it's going to be a cadence (at the G7) but then isn't (at F/A).
The cadential progression, here, functions as an elaboration of the Dm harmony, not as a structural cadence, since it then moves on to a G7. Tonicizations are visits to key-areas or regions within a progression, while modulations are structural; so your progression starts with C, and modulates to F (since that's the key that the cadence is in, at the end of the progression), and tonicizes Dm in the middle (since it uses harmonic functions from that key in the middle of the phrase).
I'm not sure if you understand what a contrapuntal cadence actually is. It's simply using bass motion by step instead of leaping, at a cadential point; it doesn't really have the same strength as a perfect authentic cadence does, so composers use it where they don't want to have too strong a cadence. You wouldn't really use it to confirm a modulation; rather, you'd have contrapuntal cadences in an already established key, or where you don't want to fully confirm the key.
No. You could have a cadential progression where ti is implied if the harmony were full, but is absent from the texture.