Where these setups look to fall down to me is not the core, which is very clever, but the slotted housing you use with them? The 15mm slot is not a nice entrance, less than ideal shape and compromised x-sectional area?
I have only seen 2 tests with completely opposite results, so for now i'm waiting to be convinced...especially as it's a fair investment to go down this path.
Cross section area of the slot is fine. 18mm (0.71") by 300mm (11.8") (guessing on slot length since total length of the cooler is stated at 410mm) is 540 sq/cm (8.38 sq/in) which is more than the cross section area of a 3.25" exhaust tube and there are 4 of them in this design. The shape of the slot(s) also contribute(s) to the distribution equalization. The edge shape may not be ideal since it is fairly sharp but everything is a compromise.
Still compromised compared to a nice open plenum is my point. As i said, waiting to be convinced and currently i'm not by some paper specs. Real world results i have seen have been mixed and the independent result was poor?
The theory is good but useful real world only if the application is also good. Maybe they are sensitive to design parameters or fabrication tolerances?