The reduction in torque requirements has to do with the coefficient of friction of the assembly paste. Engineers usually employ this (grossly simplified) equation to come up with the application torque:T = P x d x K
where: T = torque in inch-pounds (divide by 12 for foot pounds)
P = clamp force in pounds
d = nominal diameter of the fastener in inches
K = coefficient of friction (typically about .20 for common oils or greases - .15 for extreme pressure lubes like "never-seize" or high moly / soft metallic compounds)
The "P" for a given fastener is developed by its "stressed area" (the minimum cross section) muliplied by what sort of allowable tension can be safely placed on the material it is made out of. For example - the material for head bolts is usually stressed much lower than the fasteners in the rod caps (to keep the rod fasteners light - smaller - they stress them higher). Because the head bolts are clamped to a lower applied stress there isn't a serious down-side to re-using them. Not the case with the rod fasteners - they get "used up" by the combination of high application preload and operational loading. Safe re-use isn't always in the cards for them.
There is a clever demonstration I've seen put on by fastener vendors - they put a 3/4 capscrew through a "center hole" hydraulic jack, secure both the cap and nut against a heavy washer on both sides and then put a pressure gage on the jack port. Then they find a beefy mechanic and have him torque the fastener (he can't see pressure gage) with a mechanical torque wrench to the same applied torque - with and without the special thread lubricant. The difference in pressure directly corresponds to the difference in the coefficient of friction.