I don't mean for this to become circular, but here goes:
For rigid bodies, any distribution of (discrete) forces can be equivalently expressed as a resultant force acting through the center of mass, and a moment, also acting at the center of mass. This gives only the equivalent net force and torque on the body; stress distributions through the body are NOT equivalent, of course.
Alternatively, a “resultant” force could be shown acting at an appropriate distance from the center of mass in order to include the effects of the above-mentioned moment, which then disappears. I think that this is the configuration you ended up with. There is no reason that this resultant force might not be outside of the confines of the rigid body; it simply depends on the original force vectors.