Let me give some comments.
(i)Neoclassical GE still can be a benchmark of resource allocation, since NIE has not complete its standard paradigm and united theoretical system. Uncuriously, Dingding agrees with Cheung on the definition of efficiency. 
(ii)To be ture, there is no any new implications or models in this article for us today. Two cases in this article both result from Kreps 1990 indirectly or even directly, see my notes for Young's Workshop 11. For example, that Dingding's introduction of "institutional analysis with historical-specificity" is consist with Kreps's Focus Point Principle, also see Schelling 1960.
(iii)It's not because economists donn't want to modeling transaction costs into the general equilibrium theory, but we cannot reach in near future. Also, Arrow has try his best to do that. Wait and see.