Knowledge of the mechanisms behind X doesn't always precede discovery of X (why would anyone think it would always do so?) Lack of a known mechanism was why John Newlands version of the periodic table was greeted with contempt, and Mendeleev didn't find it easy sledding at first, either. But the contempt was unjustified. Birds were able to fly before the airfoil was discovered or airflow understood. Electricity was a broad field of study (Maxwell's equations, etc) long before the electron was discovered. The how usually trails the what in the history of science.
I won't be shocked if the Emdrive is a wrong turn, but the idea we could know that in advance is sciencism - the satanic inversion of science - not empirical science.