No, you're missing my point by saying "you have to have the data," and it appears that you are missing some fundamental concepts in normativity.
You do not need data to make a moral argument based upon principle. There is no conjecture involved. (You could read some Kant on this, if you like.) You need data only to determine a particular consequentialist ethics that involves a balance of good and bad outcomes often understood in lay terms as "effectiveness." I think most commenters here would agree that effectiveness is the wrong way to evaluate this system, since epistemologically we cannot know its effectiveness. (... Although many doubt it could ever be truly effective in the presence of creative adversaries.) We can only know that is it categorically wrong.
edit: In the end, it is reasonable to assume we do not have adequate data on the successes of these programs. 2 paths then arise. 1) Then you wait for more data. 2) You examine those people that you reasonably know to have access and take a look at their moral history and actions to determine the ethical issues; it's the only somewhat good data you can add to the set to maybe get a conclusion.