Still, even the best strcmp is going to involve some byte checking unless the string is exactly divisible into large words. Wouldn't the efficiency between the

Still, even the best strcmp is going to involve some byte checking unless the string is exactly divisible into large words. Wouldn't the efficiency between the two methods mainly depend on the relative odds of non-keywords occurring?