I think that what we're seeing is evidence that humans, in general, are not capable of securely delivering the kinds of online services that they are trying to deliver. It's just too complicated, and while defenses have to be perfect, attacks only have to work occasionally to be worth doing.
Edit: not that we shouldn't expect best efforts, and financial liability for organizational failures. Prison maybe for clear proven negligence or intentional sabotage, but for mistakes? Nobody will write software anymore. When is the last time you wrote even a screenful of code without a mistake?
So we should start treating them like licensed engineers... Actually I agree with this.
And the side benefit is that we could summarily execute one every once in a while for failing to write secure code.