Oh dear we are about to get away from science and into engineering with this now.
Following along with the posts above, the pushies are both forward and aft, the reaction to these pushes initial are against engine structures which react by deforming ever so slightly before fighting it out as to which way they are going to push on the engine mounts (hopefully in the forward direction if you want to go forward).
One of the biggest forward pushes is the blades to the rotor disk to the shaft and thence to the ball bearing or thrust bearing. But leaving that single bearing to take all that load at the highest thurst condition while at the same time it might even end up with a reverse load in the other direction at idle is not agood thing for the bearing. It would rattle around at the low load condition and wear out quite rapidly. So good engineering steps in and decides that one must always have a positive moderate load on that bearing regardless of the power or engine thrust condition, so the engineers move some of the presures arround inside the machine (bleeds that go from the back of the compressor to anywhere in the innards of the engine that they see fit) and presto they can actually reverse the thrust load on the bearing to push aft in all operating conditions. So with this being the case the thrust load that gets to the mounts has got to be recovered from the momentum of the air through another structure other than the bearing.
So the only time you can really predict which way the rotor is pushing is if you break the shaft between the bearings.