restructure examples v2

[matthiasdiener]

Continuation of #950, without the more contentious dependency on #948.

• No more -lazy symlinks
• No more -mpi in name (the only non-MPI example, wave.py has been renamed to wave-nompi.py).
• Test accuracy of all examples, comparing eager/lazy/numpy, as part of run_examples.sh

Questions for the review:

• [ ] Is the scope and purpose of the PR clear?
  • [ ] The PR should have a description.
  • [ ] The PR should have a guide if needed (e.g., an ordering).
• [ ] Is every top-level method and class documented? Are things that should be documented actually so?
• [ ] Is the interface understandable? (I.e. can someone figure out what stuff does?) Is it well-defined?
• [ ] Does the implementation do what the docstring claims?
• [ ] Is everything that is implemented covered by tests?
• [ ] Do you see any immediate risks or performance disadvantages with the design? Example: what do interface normals attach to?

[tulioricci]

All examples will be running with lazy? I am afraid that the compilation time with ESDG and NS coupling will make CI take a few extra hours.

[matthiasdiener]

All examples will be running with lazy? I am afraid that the compilation time with ESDG and NS coupling will make CI take a few extra hours.

If it turns out to be an issue, we could manually blacklist those long-running examples inside (e.g.) run_examples.sh.

[MTCam]

All examples will be running with lazy? I am afraid that the compilation time with ESDG and NS coupling will make CI take a few extra hours.

We might need a mechanism to allow black-listing of certain examples from being run in certain ways. Not sure. Currently, the CI time isn't terrible. pytest still takes longer than the examples. Running all the examples in lazy isn't really new, but having no way to control which ones are run in lazy is new.