Push/pull (to me) are terms for the relative position of a fan to radiator and the movement of air through that radiator. If a fan is attached to a rad and the air flows through the rad toward the fan, it's in pull. If the air moves through the radiator away from the fan, it's in push. Intake/exhaust are also relative terms, but it's rather obvious that intake is into-the-case and exhaust is out-of-the-case. If a fan is not directly attached to a radiator (be it the H55 rad or the L9 fin stack), I much prefer the intake/exhaust nomenclature to reduce ambiguity.
The document's "standard orientation" for the 120mm AIO lists the back fan in push and the rad fan in pull. The render shows both fans facing the same direction, as intakes. So the rad fan is in pull relative to the radiator, and the back fan is in push...relative to the chassis? The reverse orientation is the same, except the case has been inverted. There are no tests with the 120mm AIO and the fans set to exhaust.
For the air coolers (is there a reason you swapped the renders' front/back positions vs the aio?), the front fan is push and the back fan is pull. The render shows the front fan as an intake, the back fan as an exhaust. So I guess the front fan is in push relative to the chassis, while the back fan (which is also affixed to the chassis) is in pull relative to the noctua heatsink?
Assuming I have that correct, isn't there concern that the 25mm intake fan will be recycling the warm air that the slim fan just exhausted? My intuition may well be wrong but there's no data to disprove it. The 48mm reverse orientation has the render with both fans as intakes for glass panels, but also the fans at the top of the case (assuming reverse orientation means inverting the entire case). The 48mm reverse, glass panel orientation performed 6C better than the 48mm standard orientation glass panels (where fans are on bottom, one intake and one exhaust)--but I don't know if that improvement came from the change is case orientation (fans on top vs bottom) or the fan direction orientation (both as intakes vs. one intake/one exhaust). I would be interested to compare direct results for an air cooled setup with:
- both case fans as intakes (this exists only for 48mm glass panels, fans on top)
- both case fans as exhaust
- one as intake/one as exhaust (this data already exists)
- the effect of case orientation on each of those three fan positions
If you've already done other testing to show that these aren't necessary, I'd love to peruse the results.
