Yup, made quick work of experimenting with a few different shapes, sizes, and positions. Of course, there are many non-controlled variables such as does the air behaves like water and what speed of the blown air would be most representative... Still, the version of the bullet that worked better in my air test indeed worked better on the water too, so there is at least some degree of correlation
Oh, and yes, I was asking myself the same questions about the flat front of the bullet not being as hydrodynamic to the draining water as a more round shape would be. It probably isn't, so a smooth protrusion inside the cover would probably work even better. But it might not - what if the sharp edge actually creates a higher local velocity for the water, thus increasing suction? I don't remember if I tried a small bullet with a round front or not, but the small bullet with the flat front worked better than the larger bullet with a round front... Anyway, lots of guess work in my setup too. And, while shaping different bullets from pink construction foam is very easy, it still takes a few minutes for each one, so it becomes tedious after you make a few...