Yes, I knew it wasn't the spin rate - if anything, they would have been theoretically heavily 'over-stabilised' in gyroscopic terms, so this was mighty puzzling at the time. Not that I was too bothered as I didn't see the cartridge in standard form as a 600 yard number and wasn't interested enough to load it with a longer range 155 SMK or suchlike.
With hindsight, I suspect the dynamic stability issue you raise about the lighter bullets was the root cause, the transition to transsonic and then subsonic speeds and accompanying turbulence destabilising the bullet. A quick look at Litz's 'form factor' numbers for equivalent flat base light thirty cal bullets suggests i7 values around 1.2 which translates to ~0.15 G7 BC. At 2,230 fps MV in the conditions of that day, that gives a calculated terminal vel of 1,060 fps at 500 and 980 at 600. (33-MOA elevation at 600 over 100 zero and 1.3-MOA lateral movement / 1 mph crosswind - not an F-Class winning combination

)
Interestingly some of the surplus east European M43 around in the UK in those days produced very high MVs of well over 2,500 fps and was claimed by those who shoot a lot of 7.62X39 generally performed better than standard velocity milspec examples even at short distances.
Ironically I'm back in this ballistic ballpark again these days having built a .30BR 'plinker' on an old Paramount 4-lug Palma rifle platform with a 1:10 twist (No significance, the barrel was almost new, spare and is a Krieger so needed a home.) With 125gn flat-base Sierra MKs this is in the essentials a slightly beefed-up 7.62X39 in internal / external ballistics terms, albeit using a tight match chamber needing thin necks and employing very high-grade small primer / flash-hole brass.