There is a difference between RISC and CISC. To sum it up in a nut shell RISC processors are good at focusing on 1 task at a time per clock cycle. CISC processors focus on chewing through a few lines of assembly code at a time and also sacrificing clock cycles at the same time. When you’ve got a relatively slow processor such as a 8mhz 286 as compared to an 8mhz 68k process trying to do too many things at once can bog your CPU down quite substantially.
The advantage of RISC especially with slow processors is that they can focus on doing one thing at a time well. Today with CPUs as they are it really doesn’t matter as a much about bogging your computer down with complex instruction sets.
With RISC performance losing traction before ARM became a thing it got to a point where Intel won out through sheer brute force of CPU speed and it became to hell with doing it the more complex way anyway because we’ve got enough CPU power to do it.
I’d say the real divergence began to occur with Intel Pentium 3 and AMD K7 chips when they began pushing well north of 1ghz and Apple was still stuck with IBM trying to produce 800mhz chips.