Of course people publishing papers are responsible for the quality of the underlying data. If the data is bad, you don’t get to pretend it’s good. You write a paper describing how it is bad.