It'd still be possible to test different films using the same developer and process. I'm not asking for a comprehensive test that covers all variables, but a test that shows how the resulting picture changes when only the film is varied would still be interesting to see. Tonal curves on their own don't tell me much, I'm afraid.