Muller's supposition that the inlets and outlets of the cowling are operating at a "vacuum" reveal just how ridiculous his thought processes are. He still hasn't tried to measure ∆T through the cowling inlet and outlets. He hasn't tried using manometers to measure ∆P across heat exchangers or the cowling as a whole. His conclusions about the effectiveness of the changes being made are nonsense³.
Then it's revealed he had the radiator fabricated to fit the space available, and since the area was less than that of the Audi car radiator, it was made deeper to match the volume. Again Muller's logic is deeply flawed.
As air passes through a radiator, it absorbs heat. But at some point, making the radiator thicker doesn't provide any further heat transfer. Without calculating the optimum core depth based on ambient and coolant temps, airflow, fins per inch, tube size, and pressure drop, any depth beyond the effective maximum just increases ∆P and drag.
Muller is convinced he has conquered the problem. He's not even close.