Simple is very good. It seems to me that the reason people were getting aquarium water back in side their topoff reseviors would be because of the cappilary action in the air tube. Because of the surface tension and hydrogen bonding in water molecules, water tends to travel along thin tubing easier. As the tubing ID gets smaller, the water can actually defy gravity and simply shoot straight up the tube. There are special pipettes used in chemistry labs that are so thin that you just touch them to a liquid source and it goes right up the sampling tube.
For us in this topoff design, when water flows from the resevior into the tank to make up the difference in volume, the air hose gets wet with tank water and the tank water goes up the tube to the level of the top of the resevior. However, because the water can travel along a thin tube easier, it can go up past the level of the resevior (despite defying gravity and pressure) and back into the resevior. Using thicker ID tubing, or creating a loop of tubing far above the height of the resevior would be the best solution to that problem