related to #972

Actually deployed synthetic journeys seem to run differently than when you test locally, in that they don't go through the same bundling step. This means that tests can pass locally, but fail when deployed, making them much more difficult to troubleshoot.

It would be great to have support in the CLI (either baked in, or at least documented) to be able to run a more comparable test locally, without needing to push semi-untested journeys.