Something doesn't seem right about the idea that the fluid near the magnet falls, while the fluid up the pipe, which is further, rises.
The upstream and the downstream both seem to be equivalent. They both perceive the same gravity, but the (assumed) rising one is 1. further from the magnet, and 2. suffers friction from the pipe walls.