What I meant was that the crypto primitives (the implementation of the math) is good enough for the application, the use/assembly of them (eg: the system) leave

What I meant was that the crypto primitives (the implementation of the math) is good enough for the application, the use/assembly of them (eg: the system) leaves much to be desired. This also only applies to the algorithms that are currently considered not weak.