You do it to every extent that you can. How many scientific theories out there, including the theory of evolution, have completely and successfully ruled out all contrary explanations for whatever observations they're designed to account for? Not many. That doesn't stop them from being viable scientific theories.
Science doesn't work that way. Successful theories are ones that lead to productive lines of research. Competitive theories are obligated to explain everything covered by the current best theory and explain additional evidence.
ID does this only in the sense that it explains everything that could possibly happen. that is not science. ID will have to predict some data points that are not yet known.