Hello Brian!
Wow, I really need a Dunce hat right about now because I diagrammed the if F then G constraint totally wrong (I diagrammed it as if F is in, G is out, and vice versa).
The only other inference that I could think of was
_ _
x/z v/w l _ _
x/z v/wWhich is if either v/w is in boat 1, then w/v is in boat 2, which would put two children on both sides. But, this cannot work if either v/w is in boat 2 because that is not the contrapositive of the constraints, therefore, we do not know what happens if v/w is boat 2. In fact, both v and w can be in boat 2 and it would not go against the rule (at least thats what I'm thinking).
I also see that if h is in boat 1, g cannot be in boat 1 because that would push f in boat 1 as well, which we cannot do. So if h is in boat 1, it is either in with 3 children or f and two children. G and F can both be in boat 1 along with two children. The opposite goes for boat 2. If h is in boat 2, then f cannot be in boat 2 because it puts g in with it which would not work. h can be in boat 2 with either three children or g and two children.
Am I close to the inference that you were talking about?