I am not yet convinced that generated data ‘poisons the well’ if there is some aspect of adversarial training. About 7 years ago when I managed a deep learning

I am not yet convinced that generated data ‘poisons the well’ if there is some aspect of adversarial training. About 7 years ago when I managed a deep learning team at Capital One, I did a simple experiment of training a GAN to generate synthetic spreadsheet data. The generated data maintained feature statistics and correlations between features. Classification models trained on synthetic data had high accuracy when tested on real data. A few people who worked for me took this idea and built an awesome system out of it. Since the poisoned well is a known thing now, it seems like a solvable problem/