When you pump water into the sac, it tries to expand, which pulls in the lengthways spring and pushes out the shortways ones. That's tension. When you release the water it reverts again.
The most important thing: the shape of the muscle is roughly accurate, I think. So we could begin mapping a skin over the top...
And another thing. Given a vertex with a known number of edges protruding from it, how close to that vertex can I push a marble of radius R before it won't fit into the gap?
If we knew that, we could roll the marble all over a model, over every crack and crevace, and connect up all its centre-positions to create a skin.