I think it's the key positivist distinction, at least in terms of social science methodology: one either describes or one explains with no possibility of reconciliation.
You can make a distinction without it being an absolute, irreconcilable distinction. It's like I can make a distinction between facts and values without it an absolute, irreconcilable distinction. I can make a distinction between provisional theories and unquestionable fact without making it an absolute, irreconcilable distinction. Engels isn't as bad as he is made to sound. You try him. Metaphysics versus dialectics.