Cartridge vs passive:
The idea is to
1. get the sensor closer to the soldering surface for a faster initial reaction time to a dip in temp
2. more closely couple the heater to the tip for faster and more efficient power coupling. I.e. out of the available power, more of it should get to the tip vs heat up the handle.
For tiny skinny pointy tips, none of this matters. When the heat has to travel through a skinny diameter of metal which is more iron than copper (if any), this is the bulk of the bottleneck. In my testing with this kind of tip, the T12 tip had zero measurable benefit over the passive. You still have to run the temp higher than normal, and you can't automatically solder to a ground plane without turning up the temp. The practical difference is zero.
For larger tips where you can actually effectively get the sensor and heater much closer to the tip, this can make a difference. In T12 vs 18, instead of capitalizing on this advantage, Hakko saves money on copper and puts half the copper and thermal mass in the T12 tip. So it ends up working barely better in some specific circumstances. Overall, it's essentially the same. They just get to charge twice a tip with half the copper and which lasts half as long.
Even with a big tip and fast response time, it takes time for heat to transfer to the joint. If this bottleneck were not so big, we should be able to practically solder at only a few degrees above the melting point of solder. This is NOT the case and it never will be. It's like when you fill your tires from an air compressor. If you want to fill your tire to 40PSI, and you set your regulator to 40PSI, you will be there for a really fucking long time. The closer the tire pressure gets to the regulator pressure, the more the air doesn't give a shit what side it's on. This is same for soldering a joint. If you want to make the joint fast, you need a significant temp differential to speed up the heat transfer. To fill your tire quickly, you may set your pressure to 80 and just be careful to not fall asleep and overfill the tire for an extra 4 minutes until it explodes.
Because of the bottle necks inherent in soldering and the inherent need to have some general but not super specific amoung of temp differential, catridge style tips have limited area to actually improve performance. If you want a to drag solder lug nuts with a BR tip at 300C, it ain't gonna happen (safely) no matter how much power you have. (And RF is gonna get you way closer than cartridge ever could). You need to up the temp differential to outpace the heatsinking because you're forcing the heat/power through the skinny tip and finite contact area. Temps, in this example, that will make the tip burn itself up. There is only so far you can go before you are just gonna have to wait for physics to happen.
If you look at the metcal recovery graph that Dave put up in his vid with the grey and red lines, you will see that for same given tip and thermal mass and power, a faster response time can increase the recovery time.... by a tiny fraction of a second. And if you are to do repeated joints of this exact size and exactly the same precise interval, you will voila... here's your improvement. Enjoy it. If you were to turn up the other iron 5 degrees, or give it 5% more power, or if you were to occasionally not solder like a robot, you might see they are now more or less effectively exactly the same. Pushing one station repeatedly exactly to its limit without a break, of course it will outpace another station that is even 1% not as capable, and over time this will become an impressive number to throw in the marketing brochure.
JBC is already pushing the boundaries by pushing enough power into the tip to reduce tip life. Part of this is because consumers want the lower heat up times in print. To increase the warm up time from cold, you need more power and less thermal mass. Many users would care more what happens after the tip is up to temp. And once the tip is to temp, that mass is working for you. It is not that impressive that a T12 3mm bevel tip can get to 300C a bit over half the time as the same tip in the 888. It literally has less than half the mass in the tip. Given they have similar power output, the better coupling thing doesn't really manifest much. The most impressive and possible only significant "performance" change between the two is the warm up time. This no doubt has a huge impact on consumer impression and sales, since this is a very visible and easily understood phenomenon. Actual practical results beyond that point? Who cares? It heats up faster. There are probably some gains achieved by fancier algorithms made possible by the faster sensor response, but this is pretty high hanging fruit and similar results could be achieved with potentially better overall performance if the stress wasn't placed on warmup speed.