Now that the CLI can run unparse performance tests, the automated test runner needs to be able to handle this. This could potentialy be done as part of 1.2.0, where 1.1.0 metrics are gathered semi-manually.