Yeah, it gets very complex. I am amazed by the Xtreme, but I note that it has a wide burner head. This means the flames can get a LOT more air to mix with to reduce the CO level, after the gas mix comes out of the burner head.
The size of the jet matters too. I haven't measured the jet on the Xtreme yet, but I suspect that it may be smallish. This would make the gas velocity out of the jet high, which would help drag in more air despite the small holes and narrow tube. With a typical small-head burner that would make the gas come out of the burner head too fast, but the larger diameter means the gas velocity is reduced to sufficiently below the flame velocity. Even so, I can see incipient signs of flame-lift-off at high power.
But there is even more to consider. Look at the Xtreme and the Snow Peak stoves. The flames come out sideways from these (well, sort of), and this means the flame takes longer to hit the bottom of the pot than with something like Pocket Rocket or the Primus EtaExpress stove. The PR and PEE shoot the flame straight up to hit the pot quickly. The greater flame path of the Xtreme and SP stoves means the combustion cycle can complete before the flame is quenched.
Can I support this claim? Yes: look at the CO level vs Clearance graph for the PEE stove by itself (in the article). Raising the pot or increasing the clearance by 5 or 10 mm can halve the CO level. Just that little extra flame path length makes the difference. Similar results obtain with the special EtaExpress pot.
However, if you put the burner head too far below the pot you start to lose peak power, and the stove vendors love to claim macho heating power, even at a cost of raising the CO emission level.
Everything is a trade-off.