Wide-ranging Safety Discussion...?

I don't pretend to be an expert, but I wanted to pull out two comments
from this and another thread:

I also think it's believable that some experienced pilots may relax their safety
margins over time. *Fine, untill the day they really need them.

and:

I cannot believe this is pilot error ,comp pilots have so much instinctive skill, that
is what allows them to deal with every thing else while there body flys the plane.
So my thoughts are medical ,or possible the main ballast dumped and the fin stayed
in and pushed the c of g beyond the recoverable range.

In my opinion, these comments, while heartfelt, point out exactly why
I think it's difficult to have any kind of meaningful safety
discussion within the community. First, of course, experienced
pilots relax (or as needed, tighten) their safety margins, that is how
one becomes an experienced pilot. Safety margins necessarily vary
according to the nature of the flight. My margins at 100 hours were
very different from my margins at 1200. My margins taking a passenger
for a ride around the airport are quite different than flying in a
competition. I think it quite simplistic to think that there is one
true set of safety margins that apply to all pilots with all
experience levels under all circumstances, and as a result, it is
impossible to "train" all pilots to a consistent level of "skill" that
will keep them 99.99% safe.

Second, I see a certain element of denial that plays into all safety
discussions. In order to be willing to continue flying (or driving,
etc.), I needed to believe that my skills, experience, and margins
were sufficient to keep me safe from making the kinds of simple errors
other (less safe) pilots made. This created a bit of a conundrum, as
periodically during my soaring career, other pilots that I readily
acknowledged had equal or greater piloting skills than myself still
managed to make fatal errors doing the sorts of things I thought I was
"safe" doing. In fact, of the ten glider pilots I've personally known
who had fatal accidents, all but two were either at my level of skill
or well beyond. There are three ways to handle this conundrum. One
is to simply denigrate the perceived skills and judgement of the pilot
post-accident (which tends to apply here on r.a.s., unless we're
talking about a well known/liked contest pilot). The second is to
find some technological solution to the problem. The third is to
reach the conclusion that NO pilot is immune from making mistakes
(particularly, the seemingly benign choices that lead unexpectedly to
a critical situation), and to adjust the way one thinks while flying
accordingly...

Marc