You can't be that smart, if you don't figure out how to bridge the "communication range". Showing that IQ is a much too narrow measurement of puzzle solving ability, not related to being smart in real world situations.
You are conflating definitions, specifically making the mistake that IQ=smart. Tabooing "IQ" and "smart" [1], we can say:
Her hypothesis is that people who have some quantified ability to solve math/logic/verbal problems, find it universally easy to communicate with people with similar ability for solving such problems, assuming you control for other factors like quantified communication ability or quantified emotional intelligence etc.
Put this way, there is no obvious internal contradiction in her hypothesis, whatever the empirical support for it might be.
Her hypothesis is that people who have some quantified ability to solve math/logic/verbal problems, find it universally easy to communicate with people with similar ability for solving such problems, assuming you control for other factors like quantified communication ability or quantified emotional intelligence etc.
Put this way, there is no obvious internal contradiction in her hypothesis, whatever the empirical support for it might be.
[1] https://rationalwiki.org/wiki/Rationalist_taboo