Kaggle certainly seems like a good route for this, making it easy for the many people who merely want Wikipedia data, who will now follow the path of least resi

Kaggle certainly seems like a good route for this, making it easy for the many people who merely want Wikipedia data, who will now follow the path of least resistance to get it. I doubt it will discourage the true large-scale bad actors for whom Wikipedia is only a tiny subset of what they are trying to download, and are sufficiently well-resourced that they can't be bothered to special-case it. It'll be interesting to see how this plays out.