But by admitting that moral questions can infact be evaluated with evidence and reason, are you not also submitting that there are right and wrong answers to moral questions?
To any particular moral problem there is a finite amount of answers, and thus a finite number of results. All of these results fall into one of three categories in the long term:
1) The net well-being has increased (Or in your words: The overall effect has been positive)
2) The net well-being has decreased (The overall effect has been negative)
3) The net well-being has not changed (The overall effect has been insignificant)
Even if we aren't capable of foreseeing the full consequences of our choices, the truth is that some choices will be better than others. Would you agree?