Hedonic adjustments are more or less at the mercy of the bureaucrats. For example, new vehicles did not go up in hedonic-adjusted price from 1997 to 2020 according to the BLS [0]. The justification is that new cars have bluetooth and other features that counter-act the price change. An automatic Toyota Corolla was $12373 brand new in 1997. It was $19700 in 2020. Our inflation numbers over those years did not incorporate a price increase AT ALL, so it's a fair assessment that real inflation was higher based on that alone. The Big Mac index has probably always been the most accurate assessment, free of political pressures.
And the most fundamental problem: the survey evaluates the median spending of a sample of Americans, but the median American spends pretty much all of their money. So you end up with the dotted line in this chart: [1]. Spoiler, it's pretty much a flat line. Real median wages don't move because CPI is calculated by measuring median wages. If prices rise, people can't <consistently> spend more than all of their money, so they buy cheaper things and buy less of them, and this changes the weights and reduces the inflation numbers. The fact that some economists with PhD's think this makes sense is why you should always do your due diligence instead of accepting a doctor's diagnosis blindly.
[0] https://fred.stlouisfed.org/series/CUUR0000SETA01 [1] https://www.bls.gov/charts/usual-weekly-earnings/usual-weekl...