Maybe you can do a screenrecording of what you're doing, so we can help you better.
From the way you describe your problems right now, it appears that you're still learning Fusion itself in addition to what it is you're trying to do. There is no such thing as 'positioning' the nodes. It's all about connecting one thing to another, but they have to be compatible things.
For example the planar tracker works with images, it tracks pixels and returns coordinates that are related to where the item you tracked is on-screen.
But the *.obj you've imported lives in Fusion's 3D world until you make it an image through the Renderer3D unless you do that the planar tracker can't 'talk' to the 3D world.
Right now your questions aren't specific enough to answer them precisely
