I didn't understand the obviousness test and now I do, thanks.
And therefore it is the name of test itself that is one of the horrific failures in this debacle.
What you described is not at all an obviousness test, it's a prior-patent test. It simply asks, has this been patented before in other ways? So it has nothing to do with whether it is obvious to someone skilled in the art. It's a red herring to think that this is testing obviosity. (ahem, new word.)
And hence when you say " maybe it's not obvious, even if in hindsight it looks simple. Maybe the solution is brilliant in its elegance and simplicity."
Right, and the answer to your Big Maybe is - unmeasurable, unreproducible, and based utterly on opinion. And therefore is a farce and incites argument from the get-go, no matter what the USPTO calls it, to try to test for it that way.
But thanks for helping me to understand why the software obviousness test has failed so badly - because it doesn't test obviousness and therefore hasn't failed. It only tests to see if it is already patented in a different form. But that makes me wonder if the test itself can be challenged, because it doesn't test at all what it implies that it tests. And if the intent is truly that obvious things should not be patentable, then the definition of the test can be proved faulty - it's obvious to me that it can