Can’t you pin a game to the 8 threads with the extra cache on the 16 core parts to get equivalent performance?

Both Windows and Linux handle asymmetric scheduling and the extra cache cores are considered performance cores.