The sizeable majority of them. These projects entail lots of earth moving and pouring lots of concrete. These are exceptionally difficult things to do in areas without easy access via railways and roads. Which also happens to be the case in mountainous terrain where most potential hydroelectric sites are located.
Even the most successful pumped hydro storage facilities are quite expensive, too. The Bath County facility cost $4 billion for 24 GWh of storage capacity [1]. Fulfilling one day's worth of electricity storage for the United States would cost $2 trillion dollars. And that's assuming that all hydro storage facilities will be as cheap as this. In practice, once the optimal sites are developed we'll have to start building hydro storage in more difficult locations. And large areas of the country without mountainous terrain are still either left without storage, or we'll have to ship large amounts of electricity across the continent and spend yet more money on HVDC lines. And lastly, this is the cost of just storing the energy. We also need to build the actual generation capacity.
These kinds of broad, sweeping statements about how much storage potential exists is analogous to people saying we can just continue to use fossil fuels but plant trees to offset carbon emissions and pointing to a study showing that the US has sufficient landmass to do this. Yes, I'm sure the math behind that kind of very high level, coarse-grained calculation is correct. But actually irrigating that much land and relocating whatever it's currently used for is prohibitively diffuclt.
1. https://en.wikipedia.org/wiki/Bath_County_Pumped_Storage_Sta...