The article mentions what the author sees as two trends behind retrocomputing (reusing the old and escaping the modern world), but I'd argue that the second one should be further split, since there are two modernities that people are running away from:
- unnecessary complexity. Since the old stuff performs the same role as the new one, minus unnecessary delays or elements introducing cognitive load, might as well use the old stuff.
- lack of control. I'm not talking about an AI takeover or shit like this. Simpler stuff: big tech has become considerably better, over time, in bossing you around so you do its bidding, even against your best interests.