Right but the distinction I'm drawing here is that, while thinking in counterfactuals can be useful, thinking in terms of *wildly implausible* counterfactuals is not. Asking questions of the form "can you imagine a situation in which immoral act A is justified" is not useful.