This seems to be using advanced macros, some sort of Church-like encoded integers, and a special error rewriter to simulate integers as types. Wouldn't this be

This seems to be using advanced macros, some sort of Church-like encoded integers, and a special error rewriter to simulate integers as types. Wouldn't this be much simpler in C++ with its value-type template parameters?