The reason for short trail on a touring fork is that long trail increases wheel flop, which becomes problematic with a heavy load over the front wheel. The larger, higher and further forward the load, the worse the wheel flop.

The reason for this is easy to see - the trail defines the distance between the contact point and the neutral balance point of the wheel; as it increases the balance point follows the curve of the wheel and rises. The greater the vertical distance, the more energy to be gained by flopping. The greater the load and the further the load is above the axis, the more energy to be gained from flopping*.

Obviously the problem goes away if we can reduce or remove the amount of energy to be gained when the wheel flops. The traditional method of achieving this was to reduce the trail to nearly zero. Since there's so much inertia with a loaded front end, the thing stayed stable at speed as long as it was loaded. That last bit's important - with no load a bike with zero trail is a beast.

Another plan of attack is to absorb the energy gain somewhere else, like a spring or a damper. This can be as simple as a piece of bungee cord tied between your downtube and the rear end of the front rack. Simple and terribly inelegant but effective. You can buy a more elegant demountable spring version of this from the likes of Rivendell.

Another way is to reduce or reverse the vertical distance to the load - the problem is worst with handle bar bags, there exist special front racks which if used correctly can position the load so it rises when the wheel flops, thus enhancing stability and rendering the whole trail argument moot. "Used correctly" means keeping the centre of mass of the load on the rack back and low.

* The actual energy gain is due to a reduction in the vertical displacement of the centre of mass of the whole bike when the wheel flops. Since the locus of the centre of mass also changes as the wheel turns, this is not as simple as I have made it sound