There are two common uses for the word "impossible" that are relevant here. Solving the halting problem is impossible in the absolute sense. We have a proof.
But think about the (mind-blowing) discovery of asymmetric encryption. Before that, many things would have reasonably described using the word "impossible" in the sense of "nobody has any idea how to do that."
Do we have a proof that "golden key" thing is impossible in an absolute sense? If so, then it needs to be explained to non-technical people in a way that conveys that. If it's just that nobody knows how, then it's not actually impossible.
I think maybe people are interpreting the article as saying "if Google and Apple wanted to, they could do it this year" and of course that's incorrect. But I read it as them just lamenting that nobody even seems to be trying.
And I think their read on the lack of motivation is accurate. The people saying it's a bad idea from a technological perspective probably also think it's bad policy -- they wouldn't trust the government with a golden key even if it were possible.
So all the Washington Post can do is try and convince people that it's good policy. And it's fine to disagree with that, but it's unfair to call them stupid for thinking that there could possibly be a solution to the problem.