D120A has summed it up neatly. My personal take on this is that it is just not worth it in terms of seat/miles/accidents especially when you look at the number of accidents where the extra protection afforded by these seats would have saved lives or significantly reduced serious injury.
High energy 'stops' such as high speed over-runs and undershoots account for a small percentage of fatal crashes, and the number where the decel was such that a rear facing pax would have survived even less.
We need to face the fact that human life is quantifiable in accountancy terms and that such things as rail and aviation safety changes will only be triggered when the cost is less than the accountants' assessment of loss of life. There is a figure which I saw quoted a while back for the 'acceptable' ratio of fatals in rail travel in terms of deaths per x thousand pax miles, and it is rare for that figure to be exceeded. Likewise in air travel, and there you have the bone-heads in insurance as described by D120A to cope with as well.