I don't think that coming up with a model that matches historical data ensures that the model is predictive in any manner. I also have trouble believing solutions that vastly simplify very complex interactions. Ultimately, I think the data is surely helpful and that there is analysis that could be done with it, but I think it's about as likely to predict the future of rent prices in SF as we are to have a model that predicts the price of stocks based on three simple variables.
That might be part of the process, but the key word here was ensure. The article seemed extremely confident that this model was predictive, to the point of suggesting very drastic actions based on the predictions, which is why I made the comment.
