The author's point is that although the author only used it for morally unambiguously good deeds, there are always malicious actors and those who believe that their side of argument is "good" (e.g. extreme left and right political ideologies).

The author believes that there should indeed be some manual review for sensitive issues like these. What if an ad buyer bought ads for AR15s or for bulk order sleeping pills? You can argue that those buyers should be caught and tried, but the harm would be already done. The author would like more oversight from Google and those alike before malicious state actors (gasp!) and/or sick individuals use this unchecked power for bad.