I think that Davis uses both sections of security register (0..63 and 64..127) for decryption, because it makes no sense to use only a static key in the 1st block of security register to be save. My opinion is that Davis uses the unique ID in the second part for encryption of a view bytes in the first part of security register.
If I empathize me with the part of Davis, I would write some coded product info (serial, revision, production date,...) + encrypted info from 2nd part of security register (unique ID) into this 1st part of security register.
But to prove this theory we need at least a second owner of a green dot datalogger which is able to read out security register to compare the data of the 0..63 byte of security register with data of rdsman. I fear that this 0..63 bytes differ for every green dot datalogger.
BR
franzz