I didn't see any statement about this problem being resolved. This is what I would do to fix it. I would insert my Fluke in the line between the front door chime screw and the front door button/new parasitic load. Test leads in "COM" and "A", Switch to AC A/mA. After any initial, power up, current, settles down, I would record the standby current. Then I would switch to min/max record and ask someone to enter the motion sensing range, I would record the max as the current that should ring the chime, and the steady state current after that, as the parasitic load while it's sending, recording, et,c. Finally, I would have them push the button, and record what I would expect to be the maximum current that should ring the chime. That's it for recording data.
So initial motion sensed or button pressed current should fire the solenoid, but no current less than that. I could perhaps use a current relay between TX HI and the button. Its N.O. contacts would go between the front door screw and TX LO. Only the 2 higher currents would pick the relay and ring the chime.
I would probably prefer to use a source resistance like they mention in their manual and insert a small bridge rectifier in line. I would have the bridge feed a little opto triac circuit, that would fire the chime. Either of these two methods could be used in an interface that they would sell you. But, all in all, I'm not too impressed with what I saw from these folks! How hard would it have been for them to lift their load long enough to allow an armature to return?