I think I get it -- some people, upon hearing that a de-worming project does 1000x more good than a local-radio-empowerment project, will simply not have the conceptual tools for understanding the statement "project A does 1000x more good than project B." Their brain literally parses statements like that as "I like project A more than project B."

In turn, and still on a conventional, non-rationalist view of things, the statement "I like project A more than project B" is the sort of thing that "reasonable" people are allowed to differ about, and each person is entitled and expected to back his or her preference by reference to some kind of identity. If Bob is an intellectual, he might like to fund education charities. If Susan is an activist, she might like to fund the radio program. If Xavier is an economist, he might like to fund whichever charity looks the most efficient, which apparently seems to be the de-worming one. Each of these people will argue that their charity is the 'best' charity, but they are not pointing to measurable qualities of the charity or to predictable consequences of backing one charity vs. another -- they are simply affirming that the charity they have chosen dovetails with the identity they have chosen.

When that uppity Xavier tells zer friends Bob and Susan that they should donate to de-worming charities because they do the most good, Bob and Susan tactfully point out Xavier's mistake by pointing out that "it's not about efficiency" -- in other words, Xavier appears (to Bob and Susan) to have accidentally confused Bob and Susan with people who identify with the subjective tribe of efficiency-promoters, as this is the only hypothesis Bob and Susan can think of that would explain why Xavier would think that Bob and Susan could possibly be persuaded that de-worming is "the best."

So, when advocating for a particular charity (or political party, or whatever) it's important to take pains to point out that the charity is not merely good at promoting the "efficiency tribe," but is also better at promoting whatever tribe the listener subscribes to than that tribe's official favorite charity.

Sample line: "De-worming does more to help people learn than building schools does. I care about education, and so I want to help people learn as much as possible. That's why I donate to de-worming projects, because it's the best way to help children learn."

Not: "De-worming is the most efficient way to help the poor."

This seems obvious and uncontroversial. Nobody claims that you should donate to whichever charity is the most efficient regardless of what they're trying to do.

It's not about how much a charity helps people, but about how much your donation to the charity helps people. This is where efficiency comes in. Efficiency measures amount of good done per $1. If you have $1000 to donate to a charity, efficiency measures how much good your donation will do.

Assuming two charities have the same goal, but one is much larger, the larger one probably helps people more. But if the smaller charity is more efficient, it is the one you should donate to because your donation would do more good there.