Forgive my ignorance, but did AlphaGo not already establish that multiplayer games of the sort mentioned in the abstract are already essentially no longer a challenge for state of the art AI players?
I'm not in a position to remedy your stated ignorance, but having skimmed TFA, I can say that your question is tangential to its theme of exploring the contexts in which AI players might best compete against one another.
"Overall, AI research has primarily placed emphasis on training of strong agents; we refer to this as the Policy Problem, which entails the search for super human-level AI performance. Despite this progress, the need for a task theory, a framework for taxonomizing, characterizing, and decomposing AI tasks has become increasingly important in recent years"
It didn't achieve superhuman levels at Starcraft or Dota, not without relying on superhuman reaction speed at least. The versions without superhuman reaction speed didn't reach the top levels.
