I'm suggesting that optimizing for officer safety at all other costs may result in more overall death (/injury) than if more emphasis were put on civilian safety as well. I very much don't have any particular data to back that assertion up in this case, but often such things are true.
Hopefully that clarifies.