I'm unhappy with current quality of 99.9% software both in Windows and Linux worlds. I just try to improve this situation by developing as reliable and secure software as I can.
This is more interesting and complex task than develop 'works for me' sort of software.
I'm lazy... so I hate to spend a lot of time debugging script X, just to find it not working because if bug in module A, used by module B, used by module C which I've finally used in my script X. To avoid/minimize such situations modules A, B and C much be reliable. (Actually, this was main reason why I've switched from developing under Windows to Linux many years ago - 3 days of debugging my Deplhi application until I've reinstalled Windows and found there no more bug now (reinstalling Delphi on day 2 won't help).)I wish spend my time thinking about right way to solve something and writing code, not for debugging. This, in turn, mean my code should be more reliable and less buggy, that's all.
"we are not already doing so": just re-read my original post, there a lot of examples. "we (both individually and as a community) will gain something tangible by doing so": see motivation above "there is a workable path to improving in this direction": just change your goals to "reliability and security must be 1st and 2nd priorities" and keep this in mind while developing