> doesn't freeze
In the sense that something that is frozen can't freeze. Apparently the melting point of sand is about 1700°C, so 1000°C sand indeed should handle about the same as 20°C sand or -100°C sand. That does sound rater neat for a large scale setup.
In a small-scale setup the convection currents, easy handling and lower insulation requirements of a water setup would probably still win out.
If water tank holds water above 70C, it’s safety requirements make it significantly more expensive. Plus water close to the boiling temperature is rather corrosive. So I cannot rule out that even for small scale sand setup will be cheaper.
Residential water heaters, at least in the USA, don't typically go above 60°C. So there's room to grow here, at least for now, without building a more exotic, and therefore expensive, system.
And the reason they don't go much higher could be that at higher temps water hardness starts depositing water stone on the heating element and lining.
[1] https://polarnightenergy.fi/technology
[2] it can boil - silicon boils at 3265°C - but it is hard to see how such a temperature can be reached