Now that you mentioned a mountain... Why again is a hole needed?

Edit: I mean that an inclined rail could be just as good as a hole, and way cheaper.

Peak-power output would be less for an inclined rail (vs the vertical shaft solution).

Yeah, but if they're a quarter as efficient and cost a tenth as much, then we might have a winner.

It may have to do with maintenance.

For the gravity-based design, you need to lower a weight directly up and down in a controlled manner (ie without swinging etc.) That basically requires a shaft in the ground. If you have a mountain and a lake handy, pumped storage would be more economical, but the advantage of this design - as I understand it - is it can be deployed in places that don't have those things.

Actually an inclined plane and a hole in the ground are identical in terms of their energy storage, both are a function of their highest and lowest points. Which suggests a very simple implementation which is to run a pair of rails up the side of a mountain and but the winder/generator at the top and tie it off to the end of a rail car filled with rocks. Sure you would give up some efficiency with the friction of rail on steel but it would be pretty cheap to implement. On some of the alluvial fans in the Mohave it would make for an interested 'excess energy' storage mechanism for solar.

That said, how are those gravity fed porch lights doing? That seems like it would be an interesting proxy here.

I have seen the 1830s-sh remains of such a system in Tasmania. They sent coal in buggies down a ~200m inclined railway to the dock below, using it as counterweight to haul goods up the hill. Win-win!

I think the important factor here is the possibility of placing these things anywhere one wants. If mountains are available, I don't see why not (except maybe they would ruin the natural beauty and hence tourism...but I don't know how much impact that has). But if they are not, you can always just dig a big hole.

Suppose you have standard rail tracks and use a surplus freight train. Congratulations, you can now store the amount of energy needed to push a single train up a single mountain. That's not nothing, but it probably isn't enough to be useful on a utility scale.

That implies that you want a weight orders of magnitude more than a train (or many trains per charge "cycle" with a stockyard at each and of the mountain to store them) You're now going to need a very tough track and wheelset. They're also going to wear quickly. Probably enough to make it impractical to build and operate.

External rails on a mountain don't allow for the optional compressed air part of the system that (according to the page) can double the efficiency of the system at less than double the cost.

Also it uses way more valuable surface area. Especially when you consider the safety measures that would be required.