I totally agree that it is much easier to pick on empirical works than on theoretical works. Everyone can find something to argue against it. Some referees can be very mean and unhelpful. I sympathize with Professor Lu if his paper was held under a different standards than other papers published in that journal.
But on the other hand, economics is science, or at least the economists are trying very hard to achieve such goal. Mathematical and statistical rigorness goes hand in hand with key observations of the reality. If the scientific methodology is wrong, then the paper is wrong. It will not be published regardless of how interesting the story is. If the problems can not be addressed right now due to the lack of data or good identification strategy, it means that the project itself is not mature enough.