There are so many bad DFT papers out there because it's cheap to do DFT compared to growing and measuring samples carefully. DFT is notoriously unreliable as a

There are so many bad DFT papers out there because it's cheap to do DFT compared to growing and measuring samples carefully. DFT is notoriously unreliable as a predictive tool in strongly correlated systems, though when electron correlations are small it works well. I mean, I want this to be true, but I put little stock in DFT that doesn't calculate observables. So yes, you're right.