Context

I noticed two patterns while implementing this epic:

• We should never run the default test name, as it takes too long even for linux runs
• The split between Linux and Non-Linux wasn't really a good one. What we really have is a split between tasks that will run all of the tests (either sync or async), and tasks that will be overridden with a TEST_NAME, which includes encryption on Linux and Non-Linux.

Definition of done

Refactor the standard tasks following the above pattern.

Pitfalls

None