But what is it about them that keeps them from sounding as good as a USA made Duncan of the same model?
Your tastes and, to a far lesser degree, manufacturing tolerances.
There's no magic pixie dust or secret sauce that someone in America can put into a pickup that a factory in Korea or China can't also get. In fact you'll find the vast,
vast majority of parts that are used to create 'American-made' guitar products are first manufactured in China/Indonesia/Vietnam/etc anyway, so regardless of where a particular product was assembled the core components are all the same. (Outside of some minor and not audio-influencing technical variations, e.g. metric threading in a Chinese product and imperial threading in an American one.)
The big difference in how Chinese or Korean pickups sound compared to American brands or boutique winds is not down to different components or
quality per se, but the recipes each use and how tight the tolerances are to match those recipes, and therefore the sounds they end up with. And which recipes sound best to you is entirely personal.
If you spec out a humbucker to have two coils wound to precisely 4.1k each with .42AWG wire with a 62x12.5x3mm AlNiCo 5 magnet, that pickup will sound the same no matter if it's pieced together by a lady in the US or a guy in China as long as the finished pickup does indeed meet those specifications exactly. It's simple physics; if the magnet is the same and the wire is the same, the end result is the same. But if the lady in the US is taking an afternoon to make the one pickup and is required to take the time to check both coils really are wound to
precisely 4.1k, while the guy in China is tasked with churning out ten pickups in the same time and is told he doesn't need to verify every coil wind is precise as long as it's roughly in the right ballpark and he's keeping the production line going, well now you're going to start to get different sounds.
Most people are familiar with how original P.A.F. humbuckers vary so much, for this same exact reason. Loose tolerances mean you'll get pickups straying far from the intended spec, and if you're expecting one sound but you receive something different, you'll probably interpret that difference as being worse since it's not what you were hoping for. Rarely do we receive something that is different than we expected or required and prefer it.
And those expectations or requirements are what mostly define which pickups we perceive as good or bad, even beyond manufacturing tolerances. If you pick up an ES-335 and expect to get a blues or jazz sound from it, but the pickups installed are actually a set of SD Distortions, you'll probably perceive those pickups as being 'bad'. If you replace those SDs with a set of Chinese underwound A2 humbuckers you'll probably perceive those cheaper pickups as being 'better', even though we know that, objectively, there is no actual difference in the mechanical quality of the pickup. It's just that the cheap PAF-clones are more suitable for your requirements and expectations than the more expensive (but inappropriate) metal machines.
To that end, other than the obvious wage disparity, the main thing you're paying for when you buy a Seymour Duncan, DiMarzio, EMG, or other higher-end pickup (let alone boutique pickups) is the time the company put into R&D for that particular design. Seymour Duncan might spend a year or two refining a new pickup recipe, while a factory in Korea might ballpark it on paper, make one prototype to confirm it's roughly in the right area, and head into mass production ASAP. That doesn't actually mean the cheaper design is 'worse'—again, the mechanical quality is likely to be identical—just that since less thought and time has been put into the design it is harder to make it suit a specific requirement. It might still sound great for someone's particular needs—'stopped clock' and all that—but it'll be harder for that person to find the pickup because nailing down what the pickup sounds like is harder when the creators weren't paying as much attention to the sound to begin with. Meanwhile Seymour Duncan can very precisely market their pickup for a particular sound, since they've refined it for a specific purpose, and so people who want that sound will have an easier time finding the SD pickup.
This is why the cheaper companies
who have spent a good bit of time on their advertising, like Irongear and Tonerider, enjoy positive reputations even though their pickups are made in the same factories as many other Chinese/Korean pickups. Because those companies have taken a little more care to work out what their design sounds like they are able to inform the public better, so people make more informed purchases and therefore are more likely to end up with a pickup that suits their needs, giving the impression that the product is technically superior. (Except, as we've established, it isn't.)
Really, people need to stop thinking of and reporting on pickups as being "bad" or "good", regardless of price and country of origin, unless of course there is an actual fundamental mechanical fault with a pickup. (Which even among the most cheaply mass-produced pickups has become extremely rare.) Outside of actual faulty products, there's no such thing as a "good" or "bad" pickup, only appropriate or inappropriate pickups. The most finely-crafted, most-expensive boutique pickup in the world might be called "bad" if it was made for one sound but tested by someone hoping for another. If someone wants a heavy metal pickup then no PAF clone, no matter what country it came from or the price, is going to sound good to them.
It's the same with any guitar, pedal or amp, too. The Gibson Custom Shop can craft a phenomenal guitar, but if you like a snappy, light tone then you'd probably think a Squier Strat is better. Line 6 Spiders aren't celebrated for their sound, but if you're a goth rock player, would you really think a '62 Bluesbreaker was better? Any product can seem 'bad' when it's used for a purpose it's not designed for by a person who wants something different and similarly any product can seem 'good' when it's used by the right person for the appropriate task.
When it comes to Duncan Designed, specifically, I am reminded of Shaun Morgan of Seether. For years he used the HB-102—the DD version of the JB—in many guitars, including his signature Schecter models, long after he could have easily afforded (or likely just been given) 'real' JBs. It was something that seemed to confuse a lot of people on the old SD forums for quite a while; why would someone who could obviously easily get the 'proper' pickup keep insisting on using the 'worse' version? Eventually some guitar magazine did a rig tour with him and he explained why he kept using the Duncan Designed instead of an actual JB: the DD version has a slightly higher output and is less cutting, so since he mostly plays high-gain rhythm it is more suitable for him than an actual JB. It wasn't that either the JB or HB-102 was 'better' than the other, it's just that one was more appropriate for his playing and that just-so-happened to be the cheaper of the two.
For the record, I agree with him; I also predominantly play rhythm and I much prefer the HB-102 to the JB. But if you're a lead player you'll likely think the JB is the 'better' of the two. It's not actually a matter of the Duncan Designed being better or worse than the 'real' Seymour Duncan, just that one recipe fits one purpose more appropriately while the other configuration is better suited to another application.