I suppose you could start with existing underwater caves, lava tubes, or abandoned mine tunnels. In fact, the tech used to dig the Channel Tunnel could be used to make enormous tunnels for this.
The biggest issue I see is that the scientists specify a depth of 700 meters for the 30 meter sphere to hold 20 MWh. No where near Germany the sea is 700 meters deep. You will likely have to go all the way to Norway to find a suitable sea - with all cable losses and maintenance issues with it.
Ah - and that's another thing: how do you maintain a generator on a depth of 700 meters?
Isn't pumped storage on land a more feasible alternative?
The ingenuity of their invention is that it makes the energy storage of a given m^3 of higher more easily. The problem with pumped storage on land is that the work done is V * rho * g * h with h the height you pump the water, V the volume of water, rho the density of water and g the gravitational constant. Storing 1m^3 of water (1Mg) at 1m gives 10MJ or 10/3600MWh, so you need to store 3600 tons of water or store 1 ton 3.6km above ground to get 10MW or some combination of both.
When you go under water the pressure increases and you have to pump against this pressure. The formula is (if I recall correctly) (P0 + rho * g * D) * V with D the depth under water and P0 atmospheric pressure. Ignoring P0, you would need a reservoir 700 meters above ground to store the same amount of energy in the same amount of water. A 700 meter water tower is harder to build than an underground reservoir.
But the tower solution doesn't need a concrete/steel pressure vessel capable of withstanding whatever the pressure is at 700 meters. Just a bladder would do.
They're in the Alps. Why not just put a bag of water uphill. Fill it using excess energy. Drain it to run a turbine.
There's a patch between Denmark and Norway which is ~700m deep. This a few hundred kms north of Germany. This is actually closer to most of Germany's wind generation than Bavaria, which is much further south.
I'm not very bullish on the technology however, maintaining anything in the ocean is a difficult challenge. My guess is a mixture of batteries and power to gas would probably wind up being easier to scale up, and cheaper in the long run.
The tech used to dig the channel tunnel would not exactly be a zero concrete solution either. I would not be surprised if free-standing underwater tubes could be built with less per storage volume, e.g. due to very different safety requirements.
Opportunistic use of caves and the like could hardly ever be worthwhile because it does not scale.
