Many people complaining about overflow checking either work in high performance computing, where it really matters how fast you can add two numbers, or in embedded systems where you can't afford either the extra space for the instructions, or the extra execution time.