Not 100% sure but I reckon the philosophy behind high noses is that you get more airflow under the car and can exploit the tea tray, floor edge, turning vanes and other appendages near the side pods. Not to mention that greater volume of air also increases the amount of air that goes under the floor. Low noses would limit the amount of air that can go under the car, unless they're designed as diffusers, similar to LMP cars. I think those fins under the noses of most cars this year attempt to create downforce because they channel the air under the nose.
I think that in a year or two we should have sufficient CFD tech to create, what I term virtual aerodynamics. Meaning controlling the laminar flow across the car to create virtual winglets by the interaction of vortices and said flow. I don't know if some teams already do this.
Lets picture this: If you create a great wing and then behind the wing you put a big plank of wood, your wing would be useless. To a lesser extent, that's what the nose does to the front wing.
You can also add that now a bigger portion air is redirected to the floor rather than to top of the car. In isolation lower nose have lower CoG and generate greater downforce, but as a complete package the high nose gets the advantage.
The floor is more effective(downforce vs drag) than the wings.