We know from driving tests that use of a hand held device is dangerous but by its nature it is difficult to collect good data showing to what extent mobile phones are causal in crashes. Unlike alcohol its not specifically tested for in reportable incidences. I doubt whether there is good data showing how many people are saved by mobile phone reporting so there would never be any good way of comparing the two.Listened to a podcast (so it must be true!) that the deaths due to people using phones at the wheel are outweighed by the saving of lives as people are able to use the phones to call for emergency services quicker.
That said, using a mobile phone at the wheel to report an emergency is not illegal so "net benefit" argument can't be used to justify general use behind the wheel.