TBH I dont see how a Push/pull doubles the CFM ...

Maybe it doesn't

I distinguish between two fans with nothing between them and two fans with narrow cooler-slots between them.

A single fan on a CPU-cooler must accelerate the air from a stand-still and into air that is standing still (in the way of the new air). Moreover, the cooler slots provide "friction-like" resistance that must be overcome. The net result is that a 56CFM fan on a CPU-cooler does not actually provide 56CFM. I would not be surprised if it provided less than half that in "genuine" air flow through the cooler.

With push-pull the second fan helps move the dead air out of the way of the exhaust (by providing the push at that point). It also provides pull that creates a "partial vacuum" between the narrow blades of the CPU-Cooler that reduces the friction-like resistance. All this help adds up. To what exactly, I don't know. One would need accurate engineering airflow measurements. However, I would not be surprised if the "genuine" airflow were increased to nearly double the original amount -- remembering that the original amount might have been half the fan's rating.

Just some thoughts. Maybe someone has made good measurements and published them on line somewhere.

Anyhow, push-pull with two 56CFM fans definitely lowered the temperature vs one 56CFM fan in my experiment. I got pretty much the same temperature with two 56CFM fans as with one 110CFM fan. But, two 56CFM were quieter than one 110CFM. At the end of the day, regardless of the actual CFM being pumped, the system operated as though two fan CFMs were additive (in my small experiment).

The above is not being sold as absolutely true in an engineering sense. Its just my "hand waving" explanation for what I observed.

Maybe it doesn't

I distinguish between two fans with nothing between them and two fans with narrow cooler-slots between them.

A single fan on a CPU-cooler must accelerate the air from a stand-still and into air that is standing still (in the way of the new air).

...the air gets sucked in from the bottom (through filters) and blown towards the HSF, then blown through and sucked out the top of the case...

My system has a similar arrangement (not two heat sinks back to back). Rather, a fan blows into the case towards the heat sink fan, then 110CFM heat sink push, then heat sink, then 110CFM heat sink pull, then a short distance 110CFM case exhaust (plus an overhead exhaust). All the 110CFM fans are actually running at about half maximum RPM (I use ASUS FanExpert to control them).

I really like the idea of refrigeration cooling -- as noted by Thee-Oddball. Even water cooling is limited by air-temperature in most cases. Refrigeration has the possibility of "going low".

Thing is, the refrigeration system has to be "powerful enough" to dissipate the energy -- or things may run hot. I considered using a small room air conditioner in combination with standard water cooling plumbing (AC unit located in basement under my office) -- would cost around $200 total with what I picked out. However, too much trouble for someone not an OC fanatic (I am fanatical about other things ).