This commentary addresses a number of problems with the benchmarks proposed for evaluating theories of short-term and working memory (Oberauer et al., 2018). First, it is shown that the proposed benchmarks intentionally exclude findings regarding the core of the working memory construct and also miss some important findings from other subdomains. For these reasons, the benchmarks cannot be considered as a valid representation of the findings on short-term and working memory. Second, it is shown that although theory-neutrality of the benchmarks was aimed for, this goal was not achieved because theory-neutrality in the formulation of the benchmarks does not guarantee inclusion of all theory-dependent findings. For these reasons, the benchmarks miss their purpose and are defined in such a way as to encourage a future theory development that studies working memory in isolation from other cognitive activities and thus misses the opportunity to stimulate a better integrative understanding of working memory in the broader context of cognition.