Yeah. The authors misinterpreted their own data. Go to figure 3. Take the best headphones at each price point (the bottom-most points in the scatter plots), and there is an obvious correlation between price and quality.
A better conclusion would be "headphones with poor audio quality exist at all price points"
A better conclusion would be "headphones with poor audio quality exist at all price points"