> So why bother?

Because you're making quite a few assumptions.

Also, I'm not too optimistic about this making it past the proof-of-concept stage due to some hard to solve and hard to scale isuess with the design but I'm perfectly ok with research institutions doing research that attempts to push the envelope. What other point would there be to their existence?

I don't mind research for research sake. But I'm just saying let's not all get over-excited about this being the next way to store energy. In practical terms battery technology makes so much more economic sense so guess where the money will be spent in production deployments at scale? Where it's economic.

I'm not to bullish on battery technology for grid level storage either.

Are you optimistic about any technology in particular for grid level storage?

Superconducting loops and hydro seem to be the most realistic options at the moment. Hydro is limited by geography, superconducting loops by technology, though there has been some real progress on the latter the last couple of years.

A combination of wind + superconducting load levelers already works quite well in that grid load fluctuations are dealt with efficiently (and in a very compact package) allowing windfarms to feed old and fickle electrical grids.

The knowledge gained there can be applied to longer load shifts but it still is a real challenge.

Sea based renewables (wave/wind technology) usually need anchoring to the floor. If the anchor is also a battery that can be recharged indefinitely, you have yourself a rather good system. The only issue I can see is sediment build up in the device.

When it is full of water, pressures are equal and drain could open from the bottom to eject water and sediment. The pressure differential could keep the valve closed when empty.