In the example code, objptr is used like this:

    // prepare new object
    if (objNew(type, xp, yp) == 0)
        // no more space, get out
        return;

    // Init some vars for snes sprite (objgetid is the current object id)
    objGetPointer(objgetid);
    newobject = &objbuffers[objptr - 1];

The documentation for objNew states, that objgetid will have the ID, after the call.
objGetPointer states the handle needs to be offset by "-1", after which objptr will have the pointer.

Now I'm a bit confused, why it's necessary to use the objptr to obtain a pointer.
Shouldn't it be the pointer already according to documentation?

And if no, what's the difference between objptr and objgetid?

The assembly code also shifts the id to obtain id * sizeof(t_obj), so it's one more reason why I'd expect objptr to be t_obj* already.

EDIT:
To be clear what I wanted to ask:
Shouldn't the example read:

    // prepare new object
    if (objNew(type, xp, yp) == 0)
        // no more space, get out
        return;

    // Init some vars for snes sprite (objgetid is the current object id)
    objGetPointer(objgetid);
    newobject = objptr;

to make any sense?

Otherwise I don't see how objgetid and objptr are in any way different.