It is mainly due to how the airfoils involved produce lift at different angles of attack in the negative angle of attack (lift) range.

A typical aerobatic pattern ship has symmetrical airfoils (whose characteristic is that the lift curve slope is a straight line (linear) from -12 deg to +12 deg) with all surfaces set at 0 degrees with respect to each other. As a result when you do an inside loop with an aerobatic airplane (positive g force toward the canopy and positive angles of attack) or an outside loop (negative g force away from the canopy and negative angles of attack) when the loops are the same diameter the g forces and angles of attack are the same.

Everything is symmetrical and unless you look for a canopy you can't tell the top from the bottom. Well sometimes the rudder and wheels give it away but the general idea is that from an aerodynamic point of view there is no difference and it flies like that is the case. It makes it really easy to fly a contest pattern.

A typical trainer ship has flat bottom airfoils with the wing set about +3 degrees with respect to the horizontal tail. This makes the airplane more of a self correcting airplane, what you need for a trainer.

The flat bottom airfoil has an interesting characteristics, that over a small range of angles of attack (-4 to +10 or thereabouts) the wing lift curve slope is linear ( it acts like the symmetrical wing). But, and it is a big but, as the negative angles of attack increase negatively beyond -4 (or so) the wing no longer acts like the symmetrical wing and for a given angle of attack has less lift.

As a result when you do an inside loop with the trainer airplane (positive g force and positive angle of attack) the airplane works like the aerobatic airplane. But, and it is the same big but, when the airplane tries an outside loop ( needing negative g force and negative angles of attack which are already offset by the incidence angle) it starts having trouble because the wing isn't producing lift as well. To keep trying to do the outside loop you push in more down elevator and eventually might run out of elevator travel before the airplane outside loops.

Bottom line, the symmetrical airfoil can't tell if it is upside or not. The flat bottom airfoil can have a lot of trouble making the negative lift to make an outside loop.

It's late and I might hope this makes sense, if not I plead diminished capacity by vitue of being too sleepy.