Cost-benefit calculations are about contingent facts, which may be different in different cases; they do not indict the very nature of activities such as charity. I too value that which is functional and creative, and I agree that simply giving people money creates harmful incentive problems, but that just means that specific charitable programs must be carefully evaluated for their actual effectiveness. Money is indeed a useful mechanism, but this doesn't mean that the default market outcome is the best possible; it would be awfully strange if deliberate altruism had no power whatsoever.

I think cost-benefit calculations usually take this kind of form. You know, "X is net bad under specific conditions A and B which usually obtain, unless C; however, ancillary considerations D, E, and F; therefore recommend Y until we get better evidence." Not: "X is bad and you're stupid for supporting it." Policy debates should not &c.