I don;t suppose anyone suggests the sodium metal and water are in the cell.
But "no danger" due to no water inside is not quite true. A substantial amount of sodium metal if submerged and there is a battery leak, that might just be a pretty serious danger, depending on how much there is, and what happened to the battery (and car) just before it went in the drink.
You might find it rather easy to exit the sinking car.... if you were still conscious after the fire and explosion.
And it is inevitable that energy density going up means the device becomes closer and closer to being a bomb. It is simple physics. You have energy storage by preventing substances which release energy when they react, from reacting. To release energy you allow them to react at a controlled rate.
If you LOSE control of the rate, a lot of the materials may react at once. That usually raises the temperature, since heat cannot get away fast enough, and temperature may increase the speed of reaction, ending up with a "boom".
No doubt there will be more details later, but certainly glass electrolyte, plated-on metals, and "all solid state battery" do not seem to describe a battery composed of liquid.
Mind, I think a liquid electrode battery is a Very Good Thing with regard to the potential for a fast recharge. Drain and replace would be the method. But the meager data provided do not seem to be indicating a liquid system so far.