The character was created specifically because when we write we use different length dashes to mean different things—subtraction through to a pause for thinking.

The automated test will have no difficulty telling them apart. Are you saying it’s hard looking at the results of this tests? You might need to use a font that makes that easy (I agree many monospaced fonts don’t, but that’s not the character’s fault. Or include a step that replaces en with 2 hyphens and em with 3.