The methodology is questionable. Experimentation would be a lot more informative. Build something with a couple different sets of technologies using the best practices for each stack (practices for a React SPA should be _very_ different from a server-rendered HTMX application). Then compare metrics like development time, code size, etc and discuss findings.