Well, RISC-V is likely easier to implement efficiently than SPARC (no register windows, etc.) and has many of the same RISC-derived advantages.

Sun had some very smart people (what is Marc Tremblay doing at Microsoft btw?), and they could also have acquired more of them, perhaps like Apple acquiring PA Semi or Qualcomm acquiring Nuvia.

Also I wonder what might have happened if OpenSPARC had happened earlier, been more open, etc. (Indeed, a main reason why RISC-V exists is that there were IP issues with other architectures.)