There is no way I can answer your question without watching your process. That is why classroom instruction is so important. Find a college that teaches these topics and take a 3D modelling class for best results.
Presspull works in context with the other solids in the area. When presspull is used on a closed boundary, it creates a solid. Otherwise, it creates a surface. If you use it on a circle without an adjacent solid, it just creates a cylinder. If the visual style is wireframe, it sometimes has trouble sensing the location of other solids. The view direction is also important as you saw in the video.
If the presspull is extended into an adjacent solid, it subtracts. If it extends outward from the surface, it adds to the solid object.
In the video, I first started by erasing all of the stuff in the way and began with your 2d profiles. I exruded the body downward and the bridge downward. Then I changed to 2D wireframe to improve selection. After moving the bridge items upward, I imprinted the polyline on the bridge and used the move gizmo to move the imprinted edges.
Then I shelled the body to create the hollow form. Only then did I presspull the circle downward through the body to create the hole.
Keep the geometry as simple as you can as long as you can. End by adding fillets and chamfers. Do that too early and everything gets too complicated.
Architect, Registered NC, VA, SC, & GA.