After reading the long thread, I just happened to look in the spam bucket, and lo! there was one from him.
And looking at it, I see why I hadn't noticed them before - I don't normally look at spam scoring this high. Chickenpox is your friend. :-) Loren Content analysis details: (29.0 points, 4.6 required) pts rule name description ---- ---------------------- ------------------------------------------------ -- 2.5 LW_BOUNDARY2 Possible ratware MIME boundary --digits 6.2 FH_FAKE_RCVD_LINE_B 1.0 LW_MULT_RECIP5 Five or more recipients in same domain 1.7 SARE_HEAD_HDR_XCSIP Message headers used which identify spam 1.0 LW_MULT_RECIP3 Three or more recipients in same domain 1.0 LW_MULT_RECIP8 Eight or more recipients in same domain 1.0 SARE_RECV_IP_FROMIP3 Received line is IP address from IP address 0.6 J_CHICKENPOX_37 BODY: 3alpha-pock-7alpha 1.0 LW_FUTURISTIC BODY: We can see the future! 1.0 LW_NONPERFORM BODY: We can't see the future! 0.6 J_CHICKENPOX_41 BODY: 4alpha-pock-1alpha 0.6 J_CHICKENPOX_14 BODY: 1alpha-pock-4alpha 0.6 J_CHICKENPOX_21 BODY: 2alpha-pock-1alpha 0.6 J_CHICKENPOX_36 BODY: 3alpha-pock-6alpha 1.0 LW_RANGE52 BODY: Reference to stock trading range 1.0 LW_1933 BODY: Reference to Securities Act 0.6 J_CHICKENPOX_13 BODY: 1alpha-pock-3alpha 0.6 J_CHICKENPOX_24 BODY: 2alpha-pock-4alpha 0.6 J_CHICKENPOX_26 BODY: 2alpha-pock-6alpha 0.6 J_CHICKENPOX_31 BODY: 3alpha-pock-1alpha 0.0 BAYES_50 BODY: Bayesian spam probability is 50 to 56% [score: 0.5008] 2.7 SORTED_RECIPS Recipient list is sorted by address 2.5 INVALID_MSGID Message-Id is not valid, according to RFC 2822