With the SARS-CoV-2 virus still circulating and evolving, there remains an outstanding question if variant-specific vaccines represent the optimal path forward, or if other strategies might be more efficacious towards providing broad protection against emerging variants. Here, we examine the efficacy of strain-specific variants of our previously reported, pan-sarbecovirus vaccine candidate, DCFHP-alum, a ferritin nanoparticle functionalized with an engineered form of the SARS-CoV-2 spike protein. In non-human primates, DCFHP-alum elicits neutralizing antibodies against all known VOCs that have emerged to date and SARS-CoV-1. During development of the DCFHP antigen, we investigated the incorporation of strain-specific mutations from the major VOCs that had emerged to date: D614G, Epsilon, Alpha, Beta, and Gamma. Here, we report the biochemical and immunological characterizations that led us to choose the ancestral Wuhan-1 sequence as the basis for the final DCFHP antigen design. Specifically, we show by size exclusion chromatography and differential scanning fluorimetry that mutations in the VOCs adversely alter the antigen's structure and stability. More importantly, we determined that DCFHP without strain-specific mutations elicits the most robust, cross-reactive response in both pseudovirus and live virus neutralization assays. Our data suggest potential limitations to the variant-chasing approach in the development of protein nanoparticle vaccines, but also have implications for other approaches including mRNA-based vaccines.