A friend function, although not a member of a class, can access its private and protected members. This provides an external function the capability to interact closely with the internals of the class.

The "do-while" loop in C++ always executes its body once before checking the condition. So even if the condition in the "while" part is "false", the body of the loop will execute once. This differentiates it from the "while" loop which checks the condition before the first execution.

The switch-case structure in C++ is best suited for scenarios where a single variable's value needs to be checked against a series of constants, often seen with enumerated values. It's a cleaner and more readable alternative to a long series of if-else statements in such cases.

Function call overhead is the extra time and resources needed when a function is called. Using inline functions can help reduce this overhead, especially in recursive algorithms, because the function's code gets inserted (or "inlined") at each point where the function is called, eliminating the need for the typical call and return process, which can be resource-intensive.

In C++, the modulo operator % is used to determine the remainder of a division operation between two numbers. For example, 7 % 3 would yield a result of 1, as when 7 is divided by 3, the remainder is 1. It's a very useful operator, especially in situations that require calculations around loops and array indices.

Abstraction involves isolating the complex reality while exposing only the necessary parts. In the context of object-oriented programming, this means defining clear interfaces for classes, which allows interactions based on these interfaces, hiding the internal complex workings. This makes the software design more understandable and manageable.

When using switch-case statements, it's essential to remember to include the "break" statement at the end of each case. Without it, the program may execute subsequent cases, leading to unintended behavior known as "fall-through." The "break" statement ensures that once a match is found, no other cases are executed.

While "int", "float", and "double" are primitive data types in C++, the "string" data type is a part of the C++ Standard Library and is, in fact, a class, not a primitive data type.

Encapsulation in C++ is the bundling together of data and the methods that operate on that data, restricting direct access to some of the object's components. This is a fundamental concept in object-oriented programming. By encapsulating data and functions into a single unit, we can hide the internal details and reduce complexity.

When dynamic memory allocation is done using new, it's necessary to release that memory using delete to prevent memory leaks in C++. This ensures that every byte of memory allocated is properly released, maintaining the health and efficiency of the system.