Great post. Depending on the azimuthal sampling volume, a function of the distance from the radar, you might well be averaging in potentially high ZDR oblate raindrops along with the debris. So you would expect a ZDR drop relative to the surrounding regions of the thunderstorm, especially the FFD but maybe also within the RFD, but maybe not getting near zero. Even legit hail cores in the northeast I think are pretty often contaminated by rain and ZDR stays higher compared with hail storms you might see on the high plains or something. Probably if you had a high LCL tornado classically formed on the tail end of an RFD and away from the rain shield you would indeed see ZDRs of 0 however so the theory is correct.