Flat storage (Dropbox), databases, graph-based organization, content addressed (git), ~~tag systems (the subject of this thread)~~, time based (apple photos).

The two I bolded are just the document/folder models. In Dropbox's case it's literally that, it isn't flat storage. In Apple photos it's just better abstracted, but it's only useful for presenting one kind of document. In either case, you can replicate this by just putting everything in one folder and sorting by time with thumbnails. The two I italicized are structured in a way that makes it even less easy to retrieve data out of them unless you become specialized in the tooling. The one I struck through (because underline is not an option???) is just metadata that can fit into anything. And, finally, graph-based organization is the one I argue is the thing that requires the most amount of upkeep. This is incomparable to just having some shortcut to another folder for "random" or "uncategorized" media. If you manipulate the content your graph may become incoherent where the connections between data, not just the data itself, need revision as well. And that's in addition to forming the links between the nodes themselves, which is the maintenance I am talking about.

I don't even see how any of the alternatives you mentioned or any of the proponents of the alternatives I've seen over the years solve the specific shortcuts to media that defies category problem you cited.

For one, it's mainly a political/economical reason like how cars are the default mode of transportation.

The problem with this analogy is there are numerous obvious alternatives to cars. Foot, bike, train, airplane, horseback, rickshaw, litter, blimp, boat, segway, helicopter... We can easily point to the alternatives. So it's easy to argue that cars are bad and that we should build out more rail (and here are the 10 irrefutable reasons why). Because the alternative can not only be plainly described, it can be both imagined and actualized. The people who are arguing that there exists an alternative to the desktop metaphor are simultaneously saying it is hard to imagine and, for all intents and purposes, it cannot be actualized either.

And we can play at this unto perpetuity. What if we say that actually the superior form of transport is none of the above, it is this different form of transport. And it doesn't involve a road, because aren't roads so confining? They prevent you from seeing more of the world. So this special form of transportation doesn't use paths or roads, but landmarks. And you can travel from landmark to landmark. But... I won't tell you how that is meant to actually be accomplished. But I will really keep arguing for this new form of transportation, and I won't show you what the alternative should be like (and if I do, it will look suspiciously like the old thing). And then I'll also say that it's bad because Alan Kay, a "techbro" as you put it (it's so disingenuous to call the Xerox Parc researchers Tech Bros, because that implies they were profit-seeking trend chasers rather than research scientists) has forevermore trapped us in the world of office work rather than the Etch-A-Sketch world of play, or whatever.

I think a big problem with the content I posted is that the people involved switch between talking about desktop as a GUI, a general interface (not just graphical), and a filing system/OS within quick succession.

Indeed, that is part of what is happening.

You wouldn't say the reason why women in video games have bikini armor is because it's the most natural way to design women characters, that everyone can understand

Of course not. But that's a shifty way to introduce the idea that everything is subject to critique (which I totally agree with), because you're implying that the desktop metaphor is as sexist and ridiculous as the bikini armor. And it can be reasonably argued that the desktop metaphor is the most natural way to design the computer interface, and it can more obviously be explained that the bikini armor is just horny men sexualizing a woman character's design.

But if everything is open to critique the way you're trying to open it, then I could just as easily argue that bikini armor is in fact the most natural way to design women characters (and I'd fail for what I think are obvious reasons), just as the proponents of the alternatives to the desktop metaphor say (but fail to demonstrate) the superior alternatives to the desktop.