Once the analog signaling mode state machine is placed into digital mode my impression is that only digital communications continue and the analog signaling becomes quiescent until the digital mode is ended and it reverts to analog again.
While I'll admit I haven't indulged in the brain-melting task of reading the spec either (I believe specs are meant for speed-reading, ha - until you begin to implement the details, then dig into it in detail), but I have implemented CCS at a high level via a "modem" device. In that, and other experience toying with 15118 over AC, it seems that the digital just exists as a secondary channel over the analog - the analog takes precedence for what it's there for (dumb, hard logic of go/no-go), and the digital is there for nuance (how to go/not go). So, both still coexist - the analog PWM and resistance feedback is there for "1000-feet view" of charging state, and the digital is there for the 10-foot view.
That is to say, while charging, the station still signals 6%-ish PWM to say "digital communication, and ready to charge", and the car still pulls the +12v side down to +6v if it's meaning to say "I'd like some of that charging, please", and releases it to +9v when it wants to indicate "I'm done/don't want". But at the same time, both of these same things are also being echoed in the digital data. In effect, same as AC J1772 (just with different PWM duty) in that regard, but with all the digital nuance taking place in the RF "noise" all around it.
I dunno, it's kinda like you're listening to an old school tube AM radio and that AM radio somehow had an HDMI port attached to it for VR goggles, or something. That's what this weird setup feels like to me, but hey, uh... at least the VR works, most of the time?