Observation is obvious to the observer.. but to an observer with a different vista.. another observation may seem strange or even in error.. i.e. the Observer problem..
Whos correct?, or even more correct?, or has blind spots?..
All three could be possible with all observers(us)..
Which means what, in the context of testing observations? That we don't really test our observations, or that there's no more basis to trust the results of the test than there was to trust the initial observation in the first place? Something else altogther? Nothing at all?