While I appreciate the intent, I have a hard time seeing how we can apply uniform standards. Furthermore, would such an approach slow down ground breaking approaches of the past be it Heckman's model of selection, the rise of quasi-experiments from Card, Angrist, Krueger, and others, and new work on partial identification being done right now. Each approach that breaks new ground might seem questionable at first.
How can you have something like generally accepted accounting principles when in micro economics we can't agree on how to cluster, how to weight, how to pick polynomial orders, or what the right counterfactual is. Perhaps the last example if the most illuminating. In most economics problems we're not debating how to add up numbers. We're simply debating which number is the best prediction for states of the world we can never see. How can we ever universally agree on that?