I don't doubt that a huge market of interchangeable parts made this easier. But it's important to separate the ends from the means here, when it comes to customer concerns. The customers that want customizability and upgradability are a vanishingly small slice of customers. (Just as are the ones who want to run Linux, and that small slice does include me.)
Can you not imagine the vast market of people that might want a laptop not loaded with OEM crap ware? Because that is exactly what could emerge if Framework manages to establish a market of commodity mobile components.
I went the basket route because it required less research for me, because I just wanted to get to my end result as quickly as possible.
Grander goals about establishing ecosystems that serve other eventual end goals is not the way that most money is spent. (Though I do spend my money that way in other areas, such as with climate action, the PC market does not matter that much to me.)
You're a potential customer of this product, your cognitive dissonance on the matter notwithstanding.
A desktop you built using standardized components you sourced from a competitive market with a plethora of alternatives specifically designed for easy assembly. Should any component fail you can obtain a replacement and perform the repair yourself. Doubtless these affordances are a part of why you chose to assemble you're machine yourself.
You do indeed care. The inability to extrapolate this to laptop machines seems obtuse.