You offer that we cannot trust our observations - except that we can trust our observations in the case of a method for validating our observations.

You see how that is rather inconsistent.

How about this instead.

We can offer, in addition to our observation, the method by which we arrived at the observation. The method of our "experiment".

We might also compare our observation with that of our "peers".

(And we might also cultivate ourselves, to make better observations. If you think such a thing is possible)

But however you slice it, the validity of the observation stands upon the authority of the observer. There's no getting around that.