There has been plenty of debate about this in other fields as well. Deirdre McCloskey and Stephen Ziliak have a particularly well-written paper titled "The Cult of Statistical Significance" on this very topic. Their main point is that statistical significance is meaningless without a discussion of magnitudes.