Hierarchically organization of most important subjects in C++ for candidates evaluation

Currently, I am responsible to evaluate new C++ candidates for my Company. After few interviews, I started wondering which would be the most "possible" fair way to evaluate candidates.

For instance, if I write down a set of questions, and setup whichever punctuation level, and I evaluate candidates according to their responses, I feel this is not enough (and, consequently, not fair) because I feel, I should consider also the weight that one specific question has in C++ "big picture". For example: a simple question about inheritance, might not have the same weight as a question related to smart pointers. Even if in both questions, the candidate shows himself as "strong", it is not the same being strong in a subject (let's say) "A", which is 40% in the overall C++ picture, than being strong in a subject "B", which is 20% in the same context.

I also know you might say: "This is industry specific question. You might require strong candidates in X point, while I might require strong ones in Y". But this is not the case.

I am thinking about building up a table of most important subjects in C++11 (not just new features but all C++ including C++11/14). I think it has no sense (for the project nature) going beyond including C++17 questions.

Thus, before building up this table I considered a better option to post this subject here a receive, please, your opinion about:

Which subjects would you consider the most important to evaluate a C++ candidate? (note: The answer to this question, should be the same to this other one: Which subjects would you consider the most important in C++?),

and

Which is the weight (%) you should assign for each one of those subjects in the overall picture of C++11/14?

Example: OOP: 40 %

STL containers: 30 %

(and so forth).