Evaluating Ephemeral Organization for Phoenix Projects Workflow

Is there any reason not to treat Phoenix projects as ephemeral units of organization? My use case is to run some code that will generate traces and use the with using_project: context manager so that it’s easy to get the resulting traces (spans = client.get_spans_dataframe(project_name=project_name) ) and then create a data set with them. Then I would delete the project to keep things tidy. Does this seem like a reasonable workflow? The goal is to programmatically generate traces, create a dataset, and run an experiment all in one go.