But beyond a certain point slowing the trains gains little safety but (eg) pushes people back to the road where (in the UK for example) we spend 10x less to save a life than on the trains.
Thus, overcaution on the trains will kill more people.
Wonder how much further you can take it. If you lose a little safety (trains) and take more people off the road, you increase the risk of a train accident. How many people go back to the roads after a train accident? Is the backlash greater (my 0 data thoughts would be that more people would return to the road then just if the train was slow). I heard somewhere that more people died through increased car accidents after 9/11 (forget time span) that would have flown but didn't.
You don't really need to propagate it all the way to actual lives lost. If there is an hour of life lost on a too safe train, that's a fraction of a life lost.
Thus, overcaution on the trains will kill more people.