Whether you care about PWM or not, it is useful information to know that the KR4 effectively uses two different drivers, at least the implementation is drastically different, and, in fact, one is likely more efficient than the other, and it isn’t necessarily the brighter of the two that is more efficient, because the emitters also play a role and have their own efficiency aspects.
I can’t answer your question directly as I have not seen either emitter’s output, but I do know that 3000Lm is only 40% brighter that 1200Lm, and anything less than 300% brighter is going to only be negligibly brighter, even if you can tell the difference in brightness, because for a light to appear twice as bright as another to human eyes, it needs to actually be 3 times brighter.
If the best tint is your goal, I can only help a little, and only by deduction and transitive law say the Nichia E21A tint is superior to the tint of Nichia 219C, generally speaking. To explain this, from what I have read here at BLF, the gold standard of tint among the tint mafia is the 219B 4500K, and this is apparently superior to any version of the 219C, again, tint wise and generally speaking. Yet the tint of E21A somehow competes with the tint of 219B. So if X is competitive with Y, and Y is better than Z, then X is better than Z, iow the transitive law applies.
I also suspect that if beam shape is another goal, then you should be made aware that a single emitter will have superior beam characteristics to triple emitters, and triple emitters will have superior beam characteristics to quad emitters. It might make more sense to compare single emitter beam characteristics, but become vastly more complex comparing 4 of one emitter to 4 of another, or it might not make sense to suggest that if one emitter has better beam characteristics than another, that anything at all can be said accurately about 4 of one emitter compared to 4 of another. I could be mistaken, but I think the beam characteristics in a quad are going to depend more on the optic than the emitters themselves.
I know none of this is what you want to hear, so I’ll just tell you what I think you really want to hear: if you put weight on brightness over tint, then get the XP-L HI V3 3A 5000K, because it is actually not a bad compromise in tint for the 4300Lm you’ll get out of it. I think maybe your attraction to Nichia 219C may have to do with an assumption (that I have seen at least a dozen times already) that the 219C tint is good because the 219B tint is good, and this simply doesn’t follow. Most tint snobs here, from what I have read and have had explained to me, aren’t all that impressed with the 219C.
That said, I personally am curious about the 219C, and from the beam shots I’ve seen, I think it is very attractive, and I intend to acquire a light with 219C no matter what the finely tuned tint tastes are of those that best know tint, and I’ll certainly choose it with a constant current driver over one burdened with PWM. Over just a handful of years considering the opinions of those that know drivers best I have gleaned that constant current drivers are more efficient than drivers with PWM, so this will help mitigate the fact that the emitters with the best tint are the least efficient emitters. Plus, it is hard not to notice that PWM kind of sucks.
I hope this helps.