It's not important how much a storage device can store. It's important how much it can recover. In this scale, one can hope to get back at best somewhere around 50% of the stored energy maximum. This is due to fact, that energy is recovered through heat engine. One place, where I can actually see the benefit of this device is flattening the demand curve for heat in conventional power stations. The biggest waste of energy happens in transitioning from one working regime to another and that could be mitigated by storing heat in place where the actual heat is generated. That instead of going through Heat>Electricity>Heat (in storage)>Electricity transformation cycle.
For wind and solar, battery storage is still the best option.

First true "breakthrough" battery technology. Kudos to CCT, also for such rapid commercialisation. Kaido's comment may have merit but I doubt an efficiency of only 50%.

"In the case of an outage, each TED device can remain active for about 48 hours."
So what they are describing is a system designed to deliver all the stored energy over a 48 hour period, or, they are saying that the insulation of the stored molten silicon can only keep the system viable for a maximum of 48 hours.
This is a concept that requires further innovation input to get at further improvements.

There was no mention of what the heat engine is. Most heat engines, like internal combustion engines, are notoriously inefficient, and the over-all efficiency of the device is likely to be less than that of a battery

How little we understand chemistry is evident in the statement--"In fact, it appears silicon even gets better at storing heat after each cycle." Anyone got a theory? Perhaps the silicon is getting purer in each cycle?

For grid-level energy storage, sodium-ion battery is more viable than lithium-ion.

"estimate" and "project" are not really words that utility-scale investors like to hear when considering a device that no one has yet put into operation. I bet they're lining up to be the third or fourth to buy an installation.

It's too big/heavy for a car? Sure, but what about a home (or small/medium apartment blg)? Then the efficiency issue noted by other commenters can be addressed - at least in winter - by using the "waste" heat within the home. Besides the obvious use with home solar, it will always allow for less infrastructure in "the grid" and better reliability.

We need more information. A lot of materials can store heat but the conversion to electricity is the key. Using the term heat engine is pretty vague. There is also no mention of how the heat is originally stored by melting the silicon. Electric heating elements?

Heat engine efficiencies are hard to define, it seems. The Carnot cycle is theoretically most efficient but impossible in practice; the subject is beyond my understanding but how does HE efficiency affect operation of this storage system? That it offers advantages of cost, reliability, recycling & safety likely outweighs other considerations?