They appear to be evaluating the ability to learn and apply knowledge of previous published works on imperative algorithms. But, as far as I can tell, they aren

They appear to be evaluating the ability to learn and apply knowledge of previous published works on imperative algorithms. But, as far as I can tell, they aren't testing for much ability beyond this. Where are the questions for logic, formal methods, semantics, mathematics of program construction, functional programming etc? Such theory is much more useful for designing APIs and gluing components together, something I imagine most Google engineers spend the majority time doing.