One of the reasons I think there should be a society that brings together all of those who are working in the various sciences involved in understanding emergent orders is that awareness of others' work will 1) reduce the rediscovery of laws/processes that have already been discovered, and 2) reduce the likelihood that there will be less right interpretations, given new information.
I would consider this article on creating cauliflower fractal patterns, summarized here to be an example of both issues. What conclusions would they have drawn if they had been familiar with constructal theory? It is obvious that their evaporation technique is an example of different kinds of flows interacting to create these patterns. Further, power law distributions are not random, so they have a less right interpretation of the data than they could if they had been familiar with the right data. The paper is interesting, but the interpretation coulduse some work.
Certainly part of the problem is that you can't read everything. And rarely does one read outside of one's area of expertise. But complexity science is, I think, necessarily an interdisciplinary science. We are finding these patterns, processes, and strctures everywhere in nature and society. Researchers in all fields need to familiarize themselves with what others are doing, and familiarize themselves with the work being done that is revolutionizing science.