I think that I did not mean that there is not large efficiency gains to be made just that they can not eliminate their biggest inefficiency, unlike MCF which just removes the iron core magnets. I also disagree that this fundamentally changes the point I made. The most efficient diode pumped lasers are about 10% efficient (the primary beam on NIF when still IR is about 1% efficient.) and there is speculation that this could be 17%-18% in the medium term. So there is maybe a 20x improvement available.

The swap from iron core magnets to superconducting magnets completely eliminates that energy cost and is current technology, if JET had superconducting magnets in 1996 it would have reduced its energy usage by a factor of 40. This reduction is obviously present in all next gen tokamak designs.

The majority of the improvements therefore still need to come from target yield where I believe they need to hit several hundred MJ if not higher, so a further 2-2.5 orders of magnitude needed there. (NIF can handle up to something like 50 I think in terms of the chamber max, not the fuel max).

In comparison JET is about 1 order of magnitude less than ITER and ITER is larger in terms of power than power plants will be.

I would also be remiss if I didn't take the opportunity to point ouit that 6 orders of magnitude in repeat rate are needed.