Aside from a general similarity of goals, do you really think the paper you linked makes this one non-novel? I have trouble seeing that. As far as I can tell, there's absolutely nothing comparable to tensor product representations or NECST in your link.