The CHEERS (CHEmical Enrichment RGS Sample) observations of clusters of galaxies with XMM-Newton have shown to be valuable to constrain the chemical evolution of the universe. The soft X-ray spectrum contains lines of the most abundant metals from N to Ni, which provide relatively accurate abundances that can be compared to supernova enrichment models. The accuracy of the abundances is currently limited by systematic uncertainties introduced by the available instruments and uncertainties in the modeling of the spectra, which are of the order of 20-30%. We discuss the possible gain of extending the current samples at low and high redshift. We conclude that expanding the samples would be expensive in terms of exposure time, but will not yield significantly improved results, because the current samples already reach the systematic limits. New instrumentation, like Astro-H2 and ATHENA, and improvements to the atomic databases are needed to make significant advances in this field.