On the Sci-Fi Stack Exchange there was a question about implementing self-destruct on spaceships and comparing that to scuttling on naval vessels. One person in a comment said, "They do implement a self-destruct on real spacecraft. If a launch is off of intended trajectory and is a potential danger they will order the rocket to self destruct."
At first I was suspicious of this statement because it seems to me that causing parts of a rocket to go careening off in random directions is no way at all going to increase the safety of the situation. Then I found this question here on this Stack Exchange, where the one and only answer verifies that most nations do implement self-destruct mechanisms and includes a recap on the situation with the Challenger disaster.
But now I am still wondering: why is it safer to have many small parts of the rocket come down separately than a fewer number of very large pieces? It's still the same mass overall and might just as well do more damage since it will surely fall over a larger area. This same reasoning has been used to disqualify the idea of blowing up an incoming asteroid as stated in this article which quotes Apollo astronaut Rusty Schweickart:
Another problem I can see is blowing up a large piece of rock only to create many smaller (but just as deadly) pieces of rock, doesn’t really extinguish the destructive power of an asteroid on collision course, in fact, it might increase it.
So what makes a rocket different? Are there any examples where the self-destruction of a vehicle was undoubtedly safer than the alternative?