That only applies if the simulation is implemented using actual quantum mechanics to give the appearance of quantum mechanics, AND our brains are implemented using this actual quantum mechanics.
Even in that case, if the simulation can be suspended and the brains measured with a noise floor well below thermal noise, then the copies can be good enough that no experiment from inside the simulation could ever detect the copy event occurring, and arbitrarily many copies can be made without further degradation.
That only applies if the simulation is implemented using actual quantum mechanics to give the appearance of quantum mechanics, AND our brains are implemented using this actual quantum mechanics.
Even in that case, if the simulation can be suspended and the brains measured with a noise floor well below thermal noise, then the copies can be good enough that no experiment from inside the simulation could ever detect the copy event occurring, and arbitrarily many copies can be made without further degradation.