Combined cycle gas turbines operate at >60% efficiency, so that is lie for a start. No reason they couldn't take their gas input from a separate gas heat exchanger.
Presumably would drop the efficiency though? Overall I suspect the aims are: 1.This is quite hot, hotter than most power sources, so the dynamics may differ. 2. Responsiveness, for a storage rather than supply solution you don't want a ramp-up time. 3. Photovoltaics can be smaller and possibly more easily maintained, but only having skimmed I'm not sure what size storage they're aiming for. (Got to be reasonably big to make insulating molten silicon practical one would think.) Papers like this do tend to go for quantity over quality on the justifications though, there's usually one or two core reasons and the rest are filler.
Overall doesn't seem an unreasonable idea, silicon is cheap. Lithium is nice for batteries because it's light, not primarily for efficiency or volume.