Radiometric dating relies on past isotope ratio predictions being very reliable.
However, this is not necessarily so. For example, in Uranium forensics, 235U and 238U ratios are about the same for all sources, but 234U varies and can sometimes be used to identify the source region. So, how can we be so sure that the isotope ratios used for radiometric dating are as reliable as we assume. What if the ratios of 238U and 235U we observe are a signature of the earth, like U234 is a signature of geographical regions?
Also, for meteorites, scientists say that the radiometric dating shows when the rock was last melted. Well, what if it was melted more than once? Does each melting reset the isotope ratios in the rock? If a rock of 4 billion years was fully melted, what happens to the isotope ratios in the rock? In pages that explain radiometric dating, that I have read, don't even mention this scenario - or the following one.
And what about the assumption that we can predict the isotopic ratios of the solar system 4 billion years ago? how is the assumption justified?