I think you mean by analogy but if a measuring device were just ten percent off the more measures made the greater the cumulative error.
So a survey of my property with a ten percent at each measure would be wrong, the boundaries with my neighbors would be wrong, my tax bill would be wrong, a measure of my net worth be wrong. And I wouldn't know by how much.
But if it were a simple matter of correcting the error it could be done. However it must be known that was error and how much and how to rescale it.
In selling some property I discovered I owned five acres of land that does not exist. Why? No one knows and no one knows how to go back and correct it.
If an “out of Africa 200'000 years ago” story is based upon genetic dating and the genetic dating method is unreliable what happens to the story?
Consider what the conclusions about the electron would be if the measurement of its charge were 200 or 400 percent in error. Or if the means of measuring charge were incapable of really measuring charge at all.
Define unreliable. Suppose it was 400,000 years ago, or 150,000 years ago, does that change the story?
First, there are other methods of dating things a lot of the time, so the scale is not entirely arbitrary, but does have bounds. Second, does the inaccuracy affect your understanding of the relative sequencing of populating various parts of the world?
More to the point, however, refining these temporal scales does nothing to refute the underlying premise of evolutionary theory, namely random genetic variation and selection by fitness for the environment. I think almost any biologist is open to dispensing with almost anything else if compelling data that leads to a contradictory conclusion is presented AND it is shown that other contradictory data is invalid. [mere clashes of data are simply an invitation for further research to settle the issue]
I picked the story of the electron because it is real, and it took a significant amount of experimental work by a significant number of researchers to refine techniques to the point where measurements were reasonably repeatable and reasonably in agreement. At no point did a discovery of measurements errors (which started off being fairly substantial) "disprove" the hypothesis of a universal quantized charge for the electron.