I agree with these points, and with Holden's view that Singer's child-in-a-pond argument is generally underappreciated rather than overappreciated.

The part that I was highlighting as misleading is the implicitly claimed effect size. People who aspire to do the most good may be connotatively misled by into thinking that alleviating global poverty is a more promising target for optimal philanthropy than it actually is.