I am not in academia any longer, so I don't have access to the full paper, but I would be really curious about the boundary conditions they see. They claim to have a region of isolated turbulent mixing, however, this would mean that they no longer have a no-slip boundary layer, something that I find hard to believe in a fully fluid system like this. Instead, I imagine the best they could likely do is have a boundary region over which the Reynolds number decreases rapidly as you move from the turbulent region to the non-turbulent region.
As an aside, this reminds me of some really cool research that a friend of mine did back in grad school which is kind of like the inverse of this. They created an active system in which turbulent mixing was bound to the surface of a vesicle (video)
