The tests here verify the correctness on a higher level than the unit tests in test/.
Thy are meant to catch issues in code that is difficult to unit-test (e.g., logging output),
and to catch invalid unit tests.
Because the tests are full-fledged simulations and sometimes do unusual things,
they can also serve as additional examples, but that is not the primary focus here.

Because results are not meant to change at all, most if not all tests are regression tests.
