Well, yes, because what you almost always actually want is the mathematically accurate result. Failing that, though, I can think of a few cases where you would want overflow to saturate (hardware outputs, mostly); I haven't thought of any cases yet where wrapping would be remotely useful.
Depending on the performance penalty you're willing to take, there's good alternatives to wrapping, such as: A) 64-bit integers, B) bigints, C) crashing on overflow.
Sure, saturation might be useful in certain circumstances, like manipulating audio samples, but it's generally something you want to make explicit. Using it everywhere? It's madness!
With saturated arithmetic, you could add the (unsigned) sizes together without a possibility of an overflow, so you could eliminate all except the last range check (=branch).
If the end result is larger than what is reasonable, return an error. It's not possible that the value wrapped around and later code will be writing to unallocated memory.
Uh, what hardware are you talking about? x86 and ARM at the least have saturating arithmetic instructions.