i suppose obviousness could be 'tested' for if the end result of the patent can be deduced by only looking at the end result and not reading any of the patent filing.
Yes, I've had this thought too. Have a panel of engineers who get to see a description of the problem the invention supposedly solves, but get no information about how it solves it. If any of them come up with the same idea within a couple of days, it was obvious.