A plethora of times I've seen words to the effect of, "IQ isn't necessarily an accurate indicator of intelligence" (although that particular quote is far less egregious than I've seen), implying that I.Q. is often inaccurate in measuring intelligence. Rather than individually addressing instances of said claims, I'll expound upon the topic here in a holistic way.
Generally, I.Q. tests are a combination of measuring: pattern recognition, verbal comprehension, mathematical ability, vocabulary and short-term memory. Whilst these are not perfect measurements of intelligence, they are heavily correlated with predictions of how well peers will rate a person's intelligence, school and workplace abilities: (https://onlinelibrary.wiley.com/doi/abs/10.1002/per.799 ; https://www.tandfonline.com/doi/abs/10.1080/00221325.1979.10533422 ; http://www.ingentaconnect.com/content/sbp/sbp/1977/00000005/00000001/art00016). Unless people's perception of intelligence is wild guessing, it matches measurements of I.Q. exceedingly well.
In fact, I.Q. predicts income and educational achievement (things which, I hope we can agree, are indicators of intelligence) better than parental socioeconomic status (http://www.emilkirkegaard.dk/en/wp-content/uploads/Intelligence-and-socioeconomic-success-A-meta-analytic-review-of-longitudinal-research.pdf). Not only that, but I.Q. is the best predictor of educational level, occupational level and income level (again, more indications of intelligence). Surprisingly, I.Q. even beats 'grades' as a predictor of educational level. The average sample size for the groups is approximately 97,000: (http://thealternativehypothesis.org/index.php/2016/04/15/the-validity-of-iq/).
Whilst there isn't a panacea to alleviate the concern of I.Q. being intelligence, there is an abundance of research to suggest that I.Q. probably measures intelligence.