Volume coupling and nearest projection mapping do not integrate well and are untested

[BenjaminRodenberg]

The issue precice/tutorials#374 basically originates from a bad integration of volume coupling into the overall code and a lack of testing. I decided to comment out the corresponding code to keep the existing surface coupling features working.

Todos

• We need some tests that ensure that volume coupling and surface coupling with nearest projection mapping don't conflict
• We need a proper integration of both features
• Add channel-transport-reaction as a pure FEniCS case to our tests