I would love to agree with you, because to me they are all obvious. But the problem lies in the test itself - it is not at all measurable. It is based entirely on opinion, and thus it varies based on the particular "expert" testimony. And what is obvious this year may not have been obvious last year, and that makes it un-pin-downable to me. To me it is a losing battle to try to strike down patent by patent on obviousness. They can go either way, and so it is a never-ending battle, and the lawyers get richer. The only realistic way to approach obviousness is to argue its fallibility, bias-proneness, and slippery-slopedness in general, and strike it down as an untennable test, which must be either replaced with an unbiased, independently-reproducible test, or the system must be revoked as a whole as unfair.
And since you are obliviously experience and open-minded about this, what do you think of that approach?