To assume, that Eminem had a Zettelkasten because he had slips and a box is the same assuming that people are just sacks full of meat. The mere presence of parts is not enough to assume that there is a whole.
You can borrow the terms from linguistics: You need cohesion for the formal wholeness of your Zettelkasten (links, separate notes, etc.) and to have a good Zettelkasten, you need coherence (the actual connections between ideas). Eminem’s box has neither cohesion nor coherence. It is almost the perfect example of what a Zettelkasten is not in the presence of its parts.
The key questions at play here are where is the work of a keeping a zettelkasten done and how is represented? Where is the coherence held? Is the coherence even represented physically? Does it cohere in the box or elsewhere?
The desk in my office (and that of countless others’) can appear to be a hodgepodge of stacks of paper and utter mess. Some might describe it as a disaster area and wonder how I manage to get any work done. However, if asked, I can pull out the exact book, article, paper, or other item required from any of the given piles. This is because internally, I can remember what all the piles represent and, within a reasonable margin of error, what is in each and almost exactly where it is at, or even if it’s filed away in another room. Others, who have no experience with my internal system would be terrifyingly lost in a morass of paper. The system represented by my desk is an extension of my mind, but one which doesn’t need to be directly labeled, classified, or indexed for it to operate properly in my life and various workflows. One could say that the loose categorization of piles is the lowest level of work I could put into the system for it to still be useful for me. However, to those on the outside, this work appears to be wholly missing as they don’t have access to the information and experiences with it that are held only in my brain.
By direct analogy, I suspect that Eminem’s zettelkasten, and that of many others, follows this same pattern. They neither require internal “cohesion nor coherence” in their systems which are direct extensions of their minds where that cohesion and coherence are stored. As far back as Andreas Stübel (1684), many (including Niklas Luhmann) have used variations of the idea “secondary memory” to describe their excerpting and note taking practices.  Many in the long tradition of ars excerpendi have created piles of slips which held immense value for them. So much so that they would account for them in their wills to give to others following their deaths. In many cases, these piles were wholly useless to their recipients because they were missing all of the context in which they were made and why. Lacking this context, they literally considered them scrap heaps and often unceremoniously disposed of them.
In the case of Niklas Luhmann’s zettelkasten, he spent the additional time and work to index and file his notes thereby making them more comprehensible and possibly of more direct use to people following his death. For his working style and needs, he surely benefited from this additional work, particularly when taken over the longer horizon of his zettelkasten’s “life” compared to others’. However, it’s not always the case that others will have those same needs. Some may only want or need to keep theirs for the length of their undergraduate or graduate school careers. Others may use them for short projects like articles or a single book. This doesn’t mean that there isn’t coherence, it may just be held in their memories for the length of time for which they need it. Those who have problems with longer term memory for things like this may be well-advised to follow Luhmann’s example, particularly when they’re working at problems for career-long spans.
In Eminem’s case, given the shape and size of his collection, which includes various sizes, types, and colors of paper and even different pen colors, it may actually be easier for him to have a closer visual relationship with his notes in terms of finding and using them. (“Yes, that’s the scrap I wrote for 8 Mile while I was at that hotel in Paris. Where is the blue envelope with the doggerel I wrote for my daughter?”) It’s also possible that for his creative needs, sifting through bits and pieces may spark additional creative work in addition to the slips of work he’s already created. Cohesion and coherence may not exist in his notes for us as distant viewers of them, but this doesn’t mean that they do not exist for him while using his box of notes.
As an even more complex example, we might look at the zettelkasten of S.D. Goitein. His has a form closer to that of the better known commonplacing practices of Robert Greene and Ryan Holiday. While Goitein had a collection of only 27,000 notes (roughly a third of Luhmann’s), he had a significantly larger written output of books and articles than Luhmann. Additionally, Goitein’s card index has been scanned and continues to circulate amongst scholars in his areas of expertise by means of physical copies rather than a digitized repository the way that Luhmann’s has over the past decade. Despite Goitein’s notes not having the same level of direct cohesion or coherence as Luhmann’s, I suspect that far more researchers are actively and profitably using Goitein’s collection today than are using Luhmann’s.
For those who are more visually inclined, an additional example of the hidden work of cohesion and coherence can be seen in the example of Victor Margolin.
In this case, Margolin is certainly actively creating both cohesion and coherence. The question is where does it reside? Certainly, like many of us, some of it resides internally in his mind and in coordination with the extension of it represented in his note cards, but as he progresses in his work, much of it goes into his larger outlines drawn out on A2 paper, and ultimately accretes into the writing that appears in the final version of his book World History of Design.
As described in his video, Margolin doesn’t appear to be utilizing his slips as lifelong tools for other potential projects, nor is he heavily indexing or categorizing them the way Luhmann and others have done. This doesn’t make his zettelkasten any less valuable to him, it only changes where the representation of the work is located.
Naturally, for those with lifelong uses of and needs for a zettelkasten, it may make more sense for them to put the work into it in such a way that it appears more cohesive and coherent to external viewers as well as for their future selves, but the variety of methods in the broader tradition, make it fairly simple for individual users to pick and choose where they’d personally like to store representations of their work. If you’re like philosopher Gilles Deleuze who said in L’Abécédaire
And everything that I learn, I learn for a particular task, and once it’s done, I immediately forget it, so that if ten years later, I have to–and this gives me great joy—if I have to get involved with something close to or directly within the same subject, I would have to start again from zero, except in certain very rare cases…
then perhaps you may wish to have better notes with the work cohered directly to, in, and between your cards? Surely Deleuze didn’t start completely from scratch each time because in reality, he had a lifetime’s worth of experience and study to draw from, but he still had to start from what he could remember and begin writing, arguing, and working from there.
This is why having a lifelong zettelkasten practice is more productive for most: it acts as a knowledge ratchet to prevent having to start from scratch by staring at a blank piece of paper. The benefit is that—based on your personal abilities and preferences—you can start somewhere simple and build from there.
Finally, I’ll mention that in Paper Machines, Markus Krajewski calls Joachim Jungius’ the “first practitioner of nonhierarchical indexing”. In talking about the idiosyncratic nature of Jungius’ zettelkasten for which “There are no aids for access, no apparatus; neither signatures nor a numbering of the cards, neither registers nor indexes, let alone referential systems that guide one to the building blocks of knowledge.” he says:
The architecture of the idiosyncratic scholar’s machine requires no mediation for, or access by, others. In dialog with the machine, an intimate communication is permitted. Only the close and confidential dialog results in the connections that lead an author to new texts. When queried by the uninitiated, the box of paper slips remains silent. It is literally a discreet/discrete machine.
If this is the case, then Marshall Mathers is surely channeling Jungius’ practices, as I suspect that many are.
Perhaps in The Merchant of Venice, Shakespeare may have just as profitably written:
Tell me where is knowledge bred?
Or in the box or in the head?