I did listen AS16 PRO, which of sensitivity to all, is a fact. 18Ω, a nice data is better than early KZ BAs iem that’s just 14Ω and 15Ω, and has very good sensitivity, pretty good impression of it.
Recently, they launched AS24 , a new multi-BA product with 24 units totally(single 12 units), designed by extreme configurations of acoustic. Multiple superposition is the way to output sound, and sensitivity on the graph will be up when two 30015 work the formant curve, but if the more BAs occupie, the nozzle will be larger. After the space increases, the position of the high-frequency formant will rebound, so it cannot be infinitely superimposed. After this adjustment, KZ treble BA is getting closer, and the midrange also let me hear the magic of 26989 through the driver and tuning, and sense of hearing is completely different from other multi-unit BA at the moment. AS24 is for those who want a full BA sound with really nice layering.
As explained, it is much easier to tune the FR with several drivers and some crossover circuit. And BA driver manufacturers have now massive catalogs with well-defined and consistent properties, making the job even easier. Compared to the knowledge and manufacturing precision necessary for stable acoustic tuning with one driver, I think the multi driver road is an easier choice.
As to quality of sound, well there is no magic formula. For starters, the FR will tend to be the main factor to convince a listener that he likes the sound. So if you know where you want to go with the FR, easier tuning and just more tuning options is an attractive choice again. Of course, if you don't a clear and subjectively effective target, then, who knows how many people will like it?
The issues with multi BA:
-It might seem counterintuitive, but as you mentioned, getting good trebles can be even more of a challenge.
-Crossover. Phase can easily be all over the place.(whether that's noticed is another story). And impedance will also tend to change a lot in a sort of chaotic way, potentially leading to massive differences in frequency response depending on the impedance of your amp. As mentioned above, FR is a big variable for listeners' preference, so this might become a bigger gamble than the manufacturer wished for. I personally remember like 3 IEMs where extra impedance made me like them more, while pretty much all the other ones with big impedance swings felt worse IMO. So it doesn't seem like a generally favorable gamble. And of course you can just get a DAP or amp with a low impedance output and forget about it. A single BA will always behave with a similar tendency, higher impedance amp meaning more treble. How much is very case dependent, but the general direction of the FR change is more treble with higher impedance output at the amp.
-typically lower minimum impedance as a bunch of stuff are put in parallel. If that goes extremely low, it could be an issue for the amp trying to drive it. Again, it's case by case, but the general trend is that lower impedance load (the IEM) means higher distortions and crosstalk for the amplifier section. You can deal with some of that using a balanced output and cable, but then the odds are that the amp's impedance will double, bringing up some concern for the FR changing with impedance.
If you know a lot and have many measurements for your gear and the IEM, you can sort of predict all that and decide when and how to compromise. I personally got fed up, and now I just avoid high impedance DAPs and stupidly low impedance IEMs(check the impedance curve over frequency, do not ever settle with one impedance number given for 1kHz!). Because let's be honest, sensitivity is basically never too low for BA IEMs. Maybe the worst had been the old ER4S because of the extra 75ohm resistors(to boost the treble) and even that was loud enough for me with most sources.
Good thing with multi BA:
When using several drivers for the same frequency range (like several drivers for the bass maybe), you can reduce distortions a little. One way is simply to have each driver actually work at a lower amplitude. Let's say everything works out well, and you get about +3dB from having 2 drivers, you can get the same listening level while each driver is a little quieter. Meaning, you can hope to go 3dB louder before reaching whatever limit exists for that driver where the distortions go well over 1%THD somewhere or whatever other sort of distortion becomes clearly noticeable (you won't see anything else measured, but of course THD isn't everything). Obviously, that's not going to make something go 20dB louder. The benefit is rather small and depending on the specific driver and your listening habits, you might never get into high distortions even with a single driver. But just the same, maybe the multidriver design will let you get the little extra dB you needed/wanted while keeping distortion under control by a hair.
Another possibility is when sticking 2 drivers against each other with one rotated 180°(both nozzles still facing the same way), so that when one blade in the BA goes up, the blade of the other driver goes down. That could help cancel out a portion of the vibrations created by those movements, and also lead to some small reduction in distortions.
I'm now late for a meeting, but I think that's mostly it. I didn't proofread, sorry for the expected mess and high level Angrish.