Thing is, even if the historical accounts were wrong about important details or even totally made up, we can probably still come up with something that fits the

Thing is, even if the historical accounts were wrong about important details or even totally made up, we can probably still come up with something that fits the description. There's the rub with experimental archaeology: you are running graduate descent against a cost function written by storytellers. Getting a match isn't enough to prove much.