I recommend going to the ESA (entomolog soc of America) and also to the ENTOMO-L-list (university of Guelph) to find the answer.

Here is a wild wild guess, type brain storming:

Studying Trichogramma species from Portugal and Spain, e.g. T. cordubensis, Vargas Piqueras, during rearing activities, the species only produced females. After the change of a honey in the adult diet, males appeared, and Stouthammer, et al. (Wageningen, NL) found out that Wolbachia was the reason of, or at least involved  in that change. The infection with these bacteria interfered in the sex ratio of the insect species. Curing the Trichogramma (first accidentally) by antibiotics contained in the honey (due to treatments of honey bees) Wolbachia disappeared and males appeared. (It is much better explained in the publications).

Now, when I went through the bibliography studied by the termite specialist António de Barros Machado, (Dundu Museum, Diamang, Angola, later Oeiras, Portugal) I found Russian literature (translated) referring that in the termite guts (don't remember the species) there were found Wolbachia. It then came to my mind that it would be very interesting to try to understand whether social insects appeared due to the extraordinary effect of Wolbachia on the sex ratio in those species, or could be a key factor.

Perhaps this could be a lead. Presence or absence of Wolbachia, on one hand, and natural or introduced antibiotics in the environment and diets, on the other. I can try to find back the references of the publications I mentioned, if you want to consider that idea. 

Evidently there is much more to the social behavior, like chemicals produced by reproductive adults, and their concentration changes along the aging of these. There is a lot of research on this subject e.g. in the social bumble bees, which present strong competition crises in the hives during the summer generation. Coexistence seems to be excluded there.