Consistent naming of benchmark scenarios

[soxofaan]

while working on #4 I struggled a bit with the current name "algorithm_invocations" (bit long and obscure) and named it "(benchmark) use case" in some parts of the implementation. But now it's even worse to have different names.

I'd propose to rename everything to "benchmark scenario" or "scenario" in short, which is more self-explanatory than what we currently have