[Symmetry]

I was just talking about the minimal viable product. Yeah, metal foam is something you can make in space but not on Earth and it might be valuable to send metal down for that reason. Using a glider will increase uncertainty about where the results end up, though, compared to a simpler measure of decent. I think accepting some ablation as the price of accuracy is probably the way to go.

[Kim_Bruning]

Depends on if you put some control software on the glider. Generally a controllable aerodynamic shape gives you a good amount of cross-range ability to pick your landing ground, and a good amount of accuracy to land there with precision.

[LargoLasskhyfv]

Honestly I don't think this would be a really big problem.

But to be absolutely sure nothing bad can ever happen(TM):

Why not have purpose built catching planes match up with them, once they are low and slow enough? I imagine something Airbus A380 sized, or large Antonovs, C5 Galaxy, releasing a clamp or hook on a cable or boom, connecting to that glider in the air. And tow it to its designated splash down area.

Like in https://en.wikipedia.org/wiki/Mid-air_retrieval

There! Solved that for you :)