First the easy question. There is nothing between the wireless base phone and port 1 and nothing between the hardwired phone and port 2. Both are connected by a separate RJ11 cord.

As to voltage, I haven't yet really measured the voltage during the disconnect. I guess that is the next step to fully verify the drop. I did measure the voltage onhook at 46.5 ( actuality BSP -47 volts). That is within limits.
Off hook is 6.6 volts which is exactly where it should be. The phone sidetone is generated between the mouthpiece and earpiece when the approximate 600 ohm load is placed across the line. One would continue to hear sidetone down to under 1 volt, although the level is greatly reduced. Since I hear absolutely nothing the voltage is absent from the line during the disconnect.

The hardwired phone on port 2 has a off-hook light the triggers under about 12 volts. Even when the port 2 phone is onhook the in-service light activates when port 1 has the disconnect, and even for about 5 second after port 1 is returned on hook. So with both phones on hook and no load on the ATA, the voltage is absent for about 5 seconds. Then dialtone returns normally. There is no indication that the ATA has unregistered.