I've not done any numbers but even at 500°c I'm not sure the stored heat would do more than supplement winter heating for a community.
An interesting exercise. Assuming ambient of 10C, heating 100 tons of sand to 500C puts something like 14 billion joules of energy into it, so quite a significant reserve. Then it becomes an issue of how much energy you can take out to balance the amount you can put in during the day - or how long you can get useful heat out of it if you don't put anything more in.
If the claim of 'months at a time' is translated to a slightly more believable 8 weeks, and it means we run the 'battery' back down to ambient, then we are talking about taking in the region of 11MW out, so enough to warm a reasonable number of homes even given some pretty large efficiency losses.
If we assume an average house needs 12KWh/ year for heating and use is concentrated into 200 days per year, then a village of 1,000 homes would need to draw around 60KWh/day. But we need to factor in efficiency losses, of which I have no idea. Let's assume the whole system is only about 25% efficient. That means our battery is going to have to be able to supply a steady 10KW, so well within the theoretical capabilities of a 100-tonne battery. And these things wouldn't need to be very big, 100 tonnes of sand is about 65 cubic metres, so the size of a small industrial unit/big shed would do it.
As you say though, this is just a reworking of the storage heater concept - which of course is itself a form of artificial geothermal energy. I suppose the advantage here is that it's pretty cheap to pile sand around heating coils rather than digging or boring holes to make ground source systems.
