So, is the moment of buoyancy stronger than the moment of gravity on the valve? Consider the length of the lever arm on each. Have you tested your theory out in the pool? I sure have.
Here's an experiment. Empty out the BC and let the scuba unit settle to the bottom. Rigid or no, the butt will be up and more so as the tank empties. Move the tank around and see how it changes the effect. In my tests with my tanks (different tanks will react differently I'm sure), I have to move the tank as low as possible to get the greatest lift on the butt. On a diver, the legs are always negative as the chest cavity provides all sorts of lift.
For the most part, when I have a leg heavy student in the pool, I check to make sure the tank is at it's lowest and then proceed to shift weights forward. My favorite is the top tank cam band. Usually, moving a weight or two there and I'm done.
it's unfortunately not quite that simple. while it is true that the lower you move the tank the higher the butt will float, what you are unable to measure is how much downward force and where that force is being applied to the floor. The rig is still negatively buoyant, so there is a net downwards force. What you see in the video below is the tank starting to pivot on the crown so the CoB is actually ahead of the crown of the tank. What that means is that the net force is still downwards at the valve regardless of what the back end was doing. Since the whole rig still sinks, the net buoyancy is still negative. By moving the tank downwards, the CoB of the rig starts to shift lower on the body.
Think about it like a pair of Seesaws and ignoring the weight of the actual mechanisms.
Seesaw 1: We will use a full AL80
You have a 4lb weight *roughly the weight of a valve and first stage* on one end that is 6" away from the pivot point. It is putting a 2lbft moment on the center of the lever. On the other side, you have a .5lb mass that is 1ft away from the pivot point putting .5lbft moment. Estimated based on the mass of the gas along the length of the tank, and assuming that the tank has a CoB of the mid point of the tank. Numbers are going to be plus or minus, but the relation is the same.
The net moment arm +1.5lb on the left which is valve down with a 4.5lb negative buoyancy push at the pivot point. That pivot point is the CoB of the tank moves towards the valve as the tank is drained. As long as the tank remains negatively buoyant, there is always negative buoyancy at the CoB of the tank. If the tank is floating, then there is a net lift at that point, however that is VERY low on gas, and not really what you trim yourself out with.
Now, the next seesaw is the diver. CoB is a bit too far towards the head because we are a bit foot heavy. If we have a net negative buoyancy from the tank, and our pivot point is a bit too close to our head, we need to shift it back to our feet. We can do that in one of two ways. We can apply something with positive buoyancy below the CoB *i.e. packing air in the feet of a drysuit*, or we can apply something with negative buoyancy above the CoB *trim weights*. Since the tank is net negatively buoyant, and we know from watching the video below that it's CoB is somewhere around the crown of the tank *whether it is restrained by cam bands or not*, we want to put that mass as far towards the head as possible in order to help balance our seesaw.
Regardless of what the first seesaw looks like, flat, tail up, tail down, as long as it is negatively buoyant, and it's CoB is towards the head of the CoB of the diver, it will push the head down. This is exaggerated in doubles where the extra mass of the manifold causes the tanks to literally rock on the crown much sooner than in singles, and because of that we try to get them as low as possible to prevent being head heavy.
In sidemount, it is exaggerated the other way when we are typically foot heavy and try to get the tank valves as high up on our bodies as we can.
