My guess is that in interpreted languages, the efficiency benefit in using switch statements is indeed smaller than in compiled languages; the only actual benefit I could think of, is that in a switch statement, the operand (the one which is compared to the different 'cases') will only be evaluated once and will be immediately kept in a register, which won't change and only compare to the different 'cases', whereas if statements could potentially miss that bit and re-evaluate (even if the cost is merely reading from the memory) the operand for every if clause.
In addition, you should also think about readability. In most cases, this performance diference is quite negligible, and you should choose the option which makes your code more readable and understandable.