The proposals to redefine four of the base units of the SI that are expected to be approved in 2018 were first published in 2006, although many of the ideas had been published previously. The motivation given in that first publication mixes specific technical arguments that highlight both advantages and disadvantages with some more general arguments. We identify these more general arguments and discuss how they have been used to strengthen the rationale for adopting the proposed changes and consider whether they have also influenced the way the arguments are presented. We suggest that amongst such arguments, the increased opportunity to perform primary realizations of the definitions of the base units greatly strengthens the case in favour of such changes and would address one of the longest-standing axioms of measurement science.