Another fun conspiracy theory.
The BLS actually publishes a CPI analysis based on a consistent series without adjustments called R-CPI-U-RS. The difference between CPI-U and R-CPI-U-RS is like 0.8% IIRC [1].
By all means, prove yourself wrong. You can dig up some old newspaper ads from the 70s. None of this is something you can't calculate at home.
> That is to say that the detailed level inputs are really good but the end product is manipulated.
All the details, data and weightings are published. Are there some you specifically disagree with?
Citation needed. And for the love of God don't cite Shadowstats lol, which is trivially and obviously wrong.
[1] https://www.bls.gov/cpi/research-series/r-cpi-u-rs-home.htm