I'm not really following this argument. Because someone released Ethernet cards with duplicate MAC addresses, we should conclude that a future hypothetical USB

I'm not really following this argument. Because someone released Ethernet cards with duplicate MAC addresses, we should conclude that a future hypothetical USB-PD spec capable of powering a kettle would be more prone to failure than the status quo of bringing your own extension cord? Electric cords don't need to self-declare their capabilities, USB-PD cables do. So the odds of a mistake on the manufacturer side are lower, as you'd need to pair an incapable wire with the wrong E-Mark chip. For an extension cord to melt and catch fire all you need to do is for the customer to misread a label, or more likely not read one at all. If we're talking about the short-term motives of sociopaths, your hypothetical factory will have an even easier time labeling headphone wires as capable of carrying 16A. They don't need to fraudulently manufacture one with the wrong E-Mark chip, just slap a new sticker on the box they're shipped in. I'm as frustrated with the confusion around USB-C as anyone, but I don't think we need to worry that the problems associated with race-to-the bottom manufacturers in the 15W space are going to extend to 100W, 240W, or hypothetical 3000W cables. The price point, serious failure modes, and potential liability will all serve as moats.