Let's pare it back a bit: There's large per-country variation in the data, and many of the more restrictive measures show significant positive effect in some countries. There is no attempt to control for how successful particular countries were in implementing the policies (which one might code describe as 'non-linear effects,' per the caveats section of the paper).

The US does particularly badly with the more restrictive measures, and has a large segment of the population that denies covid is real and/or anything to worry about. Connecting those dots isn't THAT much of a leap, in the context of an internet forum...