It appears to be about as efficient as a pumped storage hydro facility (e.g. here's one in Massachusetts, built in 1970 or so - https://en.wikipedia.org/wiki/Bear_Swamp_Hydroelectric_Power...)

A gas-based design seems like it would be better at a small scale - e.g. the facility in the link has a reservoir the better part of a mile away from the turbines, and has a max output of 600 MW or so.

CO2 may actually be a good working fluid for the purpose - cheap, non-toxic except for suffocation hazard, and liquid at room temperature at semi-reasonable pressures. I'm not an expert on that sort of thing, though.

> A gas-based design seems like it would be better at a small scale

The major advantage over pumped hydro would be you do not need very specific geography to make it happen (90 - 300+m change in elevation)