Hydrostor's technology looks good to me. In principle, it should be scalable to any degree needed. The "purpose build air caverns" are probably a limiting factor on specific cost of storage, but because they're storing compressed air that will be heated before expansion through a turbine, the storage capacity per cubic meter of cavern will be high.

As to round trip storage efficiencies, Hydrostor is only claiming 60%. That's beats the efficiency for the ~50% for McIntosh CAES plant in Alabama and thr ~40% for green hydrogen when it's used for energy storage, but it's below the 70% that studies had indicated as feasible for an adiabatic CAES facility like Hydrostor's. But I recall back in the early 1960's when a pumped hydroelectric storage facility was being built west of where I grew up in Colorado. Its round trip efficiency was only 33%. Over the years, I watched as new projects were implemented, and round trip efficiencies grew to 45%, then 60%, and then 70%. Now I understand that they're approaching 80%. I suspect that efficiencies for advanced CAES will experience similar improvement -- provided that enough new systems get built to make the improvements worthwhile.