It is, in theory (hedonic price adjustment can go both ways), but I don't know how accurate their measurement is.
Edit, now that I checked it looks like hedonic price adjustment measurements are performed on only 7.5%[1]of the goods in the CPI basket, and the main goal seems to be to avoid overestimating inflation by tracking quality improvements better.
The examples that the article give are "memory size and CPU speed for computers or horsepower and miles per gallon for cars", that is, technological improvements that would be a reason to adjust inflation value further down because "quality" went up. Without, of course, taking overall lifetime of the product.
So this would in fact make the inflation misreporting problem even worse.
Edit, now that I checked it looks like hedonic price adjustment measurements are performed on only 7.5%[1]of the goods in the CPI basket, and the main goal seems to be to avoid overestimating inflation by tracking quality improvements better.
[1]: https://www.nber.org/digest/20239/correcting-quality-change-...