Isn't that the kind of thinking/stats that kept the flawed Ford Pinto or the road despite being a time ticking bomb ?
No, not at all.
The Pinto *did* have a flaw that made it significantly more dangerous if it was in a typical rear-end collision. The reasoning was economical: Ford saved money by paying out enormous settlements when the flaw happened to cause greater injury or death than would normally be expected in an accident, rather than recalling and repairing ALL flawed Pintos.
The analogy, in this case, would be if the risk was already proven (which it's not; there has NEVER been a comprehensive analysis of how actual driver behavior or accident rates is affected by cell phone use), and I was suggesting that the COST of the accidents was lower than the COST of passing and enforcing legislation.
I am not suggesting anything of the kind. If we do the analysis, and find that, yes, cell phone use HAS increased the danger on the road, we absolutely *should* ban use of phones while driving (and should also look at other factors involved, as ceoyoyo mentions below). But until we've actually examined what *really* happens, rather than using inappropriate proxies in controlled conditions, we should stop assuming we know how everything works.