The problem this combination has
The base is shortbread: one part sugar, two parts butter, three parts flour, with no egg, no leavening and no liquid binder. That absence is what gives it its sandy crumb, and it is also what makes it the least water-resistant dough specified on this site.
Put a layer of cooked apple on top and two things happen at once. The fruit releases water as it bakes, and the dough underneath has no protein network to resist that water. Nothing egg-bound holding it together. Sogginess here is not a failure of execution; it is the default, and a recipe that works is one that has done something about it.
The tart doughs give the comparison in a measured form, because holding a wet filling is a recorded column in that dataset rather than an opinion.
What the doughs that cope do differently
Of the three tart doughs specified on this site, the one recorded as strong enough to hold a liquid filling without leaking is pâte sucrée, and what it has that shortbread does not is egg. The egg sets during baking into something that resists water rather than absorbing it. Pâte sablée, higher in butter and lower in sugar, is recorded as weaker and more fragile; pâte brisée, bound with cold water, is moderate.
So a shortbread base carrying wet fruit is working against its own composition, and there are only two honest routes: get the water out before it meets the dough, or bake the base before the fruit goes near it. The second is blind baking, which this site specifies with temperatures and times, and which exists for precisely this problem.
The full comparison of the three doughs on sugar, butter, binding liquid and method is at pâte sucrée and pâte sablée.
Which apple, and why the register lists two
The register carries two apples, and the difference between them is the point here. An eating apple is recorded at pH 3.30–4.00 with its pectin not established. A tart cooking apple is recorded at pH 3.33–3.33 with high pectin and set group 1, the group that needs no help to set.
That matters for a reason that has nothing to do with jam. A high-pectin, high-acid apple cooks down to a layer that holds together, because the pectin sets it. A low-pectin eating apple cooks to something looser and wetter, which is exactly the wrong thing to put on this base. Choosing the apple is choosing how much free water ends up sitting on the shortbread.
Both entries, with their sources, are in the register.
Questions about apple shortbread
Why does the shortbread base go soggy under apple?
Because shortbread has no egg and no liquid binder. That absence gives it its sandy crumb and also leaves it nothing to resist water with: the fruit releases water as it bakes and the base absorbs it.
Which tart dough holds a wet filling best?
Pâte sucrée, per the tart-dough dataset, which records it as strong enough to hold a liquid filling without leaking, and what it has that shortbread does not is egg. Pâte sablée is recorded as weaker and more fragile.
Should the base be baked before the fruit goes on?
It is one of only two honest routes, the other being to reduce the water before it meets the dough. Blind baking exists for this problem, and this site specifies it with temperatures and times.
Does the type of apple matter?
Yes, and the register lists two for that reason. A tart cooking apple is recorded with high pectin and set group 1, and cooks to a layer that holds together. An eating apple has its pectin unestablished and cooks looser and wetter, which is the wrong thing to put on a base with no binder.