I'm personally very happy that we're still funding science that isn't obviously monetised.
The reason for Antarctica is that it's the only place you find cubic kilometres of stable ice that doesn't drift around.
All of it was driven by a flood of experimental results that were not compatible with classical physics.
See "Ultra Violet Catastrophe" in black body radiation, photo effect.
Planck famously performed curve fitting on the black body spectrum and only afterwards found the theoretical foundation for the function that fit the data.
I'm actually not sure exactly what "they" is in this sentence.
I know about the Ultraviolet catastrophe, but that's cherry-picking.
Dirac predicted the Positron before it was discovered.
Quantum entanglement was predicted by Einstein, Podolsky and Rosen before it was shown ("Spooky action at a distance")
Black Holes and Gravitational waves were both predicted before they were detected.
Einstein predicted time dilation before it was detected.
Szilard predicted chain reactions in Nuclear fission before they were demonstrated.
And probably most pertinent to this discussion, Pauli predicted Neutrinos before they were detected.
Half of the list above are critical for GPS, energy, semiconductor research and many other fundamental technologies of the modern age. They were researched and predicted without any thought of a "monetisable" product.
Because of the experimental result of missing energy in the beta radiation spectrum.
Bone of these were "pure thought experiments" as you describe. Many were side effects, new predictions of theories to address other experimental results, but none of them happened in isolation.
You simply don't make predictions about the real world in "pure thought experiments".