To answer your questions:
1. Generally, it does not hurt to include the structural surface on the SWL.
If you are only running a Hydrodynamic Diffraction calculation, it does not make any difference whether the structural surface on SWL is included or not - Aqwa will consider it as non-diffracting, and ignore it for the radiation-diffraction calculation. The same can be said for any other parts of a vessel above the mean waterline. However, if you are also going to run a Hydrodynamic Response stability analysis or time domain analysis (without slow drift), then you should include the structural surface. For these analysis types, Aqwa will calculate the hydrostatic pressure over the instantaneous wetted surface, so it is more accurate to include all of the structural surfaces in this case.
The existence (or not) of the structural surface on the SWL does not affect the internal lid. Once Aqwa decides that the structural surface is non-diffracting, it does not matter if the internal lid elements are overlapping.
2. The automatic internal lid method will create lid elements inside the toroidal surface only. The moon pool will not be covered.
3. For the external lid surface, either A or B is right. You should have a small clearance between the structural surfaces and the external lid surface - any clearance is fine, as long as they are not touching (i.e. not sharing nodes in the mesh). The Aqwa Workbench editor will automatically correct the lid normal if necessary, so you don't need to worry about it. (Generally, you need to use a Surface Flip operation in DM if you do need to change the surface normals for any other reason).
I hope this helps!