Implement better handling of plugin tests (concurrency, better failure reporting, etc)

[SwooshyCueb]

At present, the testing environment doesn't have a lot of knowledge about the tests being run when running plugin tests without explicitly passing in a list of tests. As a result, any failure reports as all tests failing, and concurrency is not possible. Let's fix that.

A potential solution I've been thinking about is having the testing environment ask the test hook to enumerate the tests (it is possible to do so programmatically, so we can let the test hook define the granularity).

[alanking]

Another possibility: irods/irods#7932

This issue could also be related to / duplicate of / replacement for this issue: #53