There's a difficult - VERY difficult - trade off between safety, absolute safety and cost.  The article that Lee's quoted talks of lighter trains less able to withstand headon collisions, but cheaper to build and more cost efficient and faster to run.   The kneejerk reaction on saftey these days seems to be "if it's a risk, then spend money on fixing it".  The sentiment is laudable, but what if the money that has to be spent then makes the product so expensive that no-one buys it, and perhaps ends up buying (or having to buy) something that's less safe?