One way to interpret this is not as the author's endorsement of the other report, but as a demonstration of how fragile these happiness rankings are to perturbations in methodology / definition.

Apropos to that: I wish the author had said more about critically evaluating tweaks in methodology and definition.

(For example, he cites Blanchflower and Bryson because he prefers positive affect as a measurement of happiness – but doesn't note that Blanchflower and Bryson pool data for 2008-2017, so in terms of rankings they may be measuring something meaningful but different.)