Thoughts on a universal Launch Escape System

I'm completely unqualified to provide any serious analysis of this subject but it is my hope that someone else more qualified may pick it up and run with it...

I get the impression that a Launch Escape System (LES) on a crewed launch vehicle is not considered the most fundamental consideration of safety. By that I mean that risk analyses tend to calculate the risk of loss of the entire vehicle, and then tack on the failure rate of the LES as an extra factor. I'm sure I've seen the probability of loss of crew calculated like this a number of times.

My question is " Can the reliability of the LES, given loss of vehicle be drastically improved?"

I suspect the answer is 'yes' for a number of reasons:

-Identified Vehicle failure modes as related to LES mostly seem to assume the worst possible scenario, eg an explosion at maximum velocity. But many failure modes may be quite benign from the point of view of the LES, eg drifting out of the correct flight path or loss of adequate thrust.

-Not many LESs have been designed. A quick online search that inevitably led to Wikipedia revealed only two that have been or will be used in operational systems, the Ares-I Launch Abort System, and the Apollo LES. There must surely be great scope for iterative improvements.

-As far as I know, an LES has never been designed as an individual project with the explicit goal of saving the crew regardless of how the rocket underneath behaves.

-If great advances in crash safety can be made in motor car racing, it is reasonable to hope that some of the same principles are transferrable to launch systems.

-The next generation of fighter aircraft, such as the Eurofighter Typhoon have extremely capable ejection systems that are intelligent enough to adjust their behaviour for the situation, to maximise the likelihood of survival for the pilot, eg by dropping them down more quickly from high altitude using a drogue shute.

Suppose NASA were to make development of a new 'safe' LES that could be fitted to either Falcon 9 or EELV's, or possibly other rockets a priority, the whole method of crewed launch vehicle development could change. With such a prototype LES in hand, NASA could deliberately search for unproven launch vehicles for use in an iterative test regime. This would make a great partnership between NASA and private companies, although admittedly each organization would have almost diametrically opposed goals during the test process. NASA would be happy for catastrophic failures to occur in order to test their system under realistic conditions, while the private entity offering the launch vehicle would be trying to ensure NASA doesn't get to test their LES too thoroughly! A nice problem to have would be where the unproven launch vehicles are found to be too reliable to meet the test objectives, forcing NASA to deliberately rig the launches to induce failures - something that hasn't been done previously to my knowledge.