Can Azure ML components write to dynamic outputs?

Question

In the Azure ML SDK v2, is it possible to have a component that writes to different locations in Blob storage based on it's input?

Here is the only way I could think to arrange this (see path in the outputs variable), but it doesn't work. I don't think the ${{inputs.output_datastore}} and ${{inputs.output_path}} are accessible in the outputs definition.

save_result_component = command(
    name="save_result_component",
    display_name="Save Result Component",
    description="Saves data at input_path to output_path in output_datastore",
    inputs={
        "input_path": Input(type="uri_file"),
        "output_datastore": Input(type="string"),
        "output_path": Input(type='string')
    },
    outputs=dict(
        output_path=Output(type="uri_file",
                           mode="rw_mount",
                           path="azureml://datastores/${{inputs.output_datastore}}/paths/${{inputs.output_path}}")
    ),
    # The source folder of the component
    code='./save_result',
    command="""python save_result.py \
            --input_path ${{inputs.input_path}} --output_path ${{outputs.output_path}} \
            """,
    environment="azureml:ffm_env:4",
)