I've made a quick test not settling the 2nd GS, but keeping him for save 10 t. on Edu.
Unfortunately, 3rd Gperson was a GSpy.
A settled GS can give his 9 raw

at any research rate: this is great, 'cause we need to REX to 7 cities, but this is a disaster for the economy.
I managed to have rifling later than in my previous attempts and in addition, liberalism was at risk.
It seems that if an AI see the chance to grab the free tech it tries to do it.
Otherwise, he will never research philo, Edu much later.
The detours: i'd like to see your opinions about the usefulness of MoM and its prolonged GAge.
Conclusion: for sure, we'll take rifling from Lib.
BTW in my attempt i suffered also from an early DoW from Hammu. Better forget it. Let's try another variant.
edit: variant tried.
Straight to Lib, rifling on turn 101. But:
Everything is building slowly, even our cheap Universities. The Forge bonus applies to the output

so they arrive 1-2 turns later. And research will suffer.
In my other attempts with forges Lib arrived on turn 103-104 and the 3 main cities were with their forges, universities, airports, barracks and some Xbow for protection.
No point in having Lib 3 turns earlier if we cannot build paratroopers, but we must build forges and structures before begin to build them.
No, on my dead body: MC ASAP. Once we traded for the prerequisite techs, go on.
Mesix spent some post to convince me to settle GSs, and i acknowledge he was right.
But this no-MC-no-forge thing... no, it's wrong, plain wrong
Another fishy thing happened in one of my attempts: i've lost pyramids very early. The only variants on this were the scout and the airship moves... strange, indeed!