Cold fusion didn't work, until it did. 3D TVs didn't work, until they did. Flying cars didn't work, until they did. Satellite phones didn't work, until they did.
I feel like there's a bit of survivorship bias here. Not saying that vibe coding is completely useless, but you still have to review the code it produces in order to make the most of it really. It isn't completely autonomous (yet) to the point where it can scale to any of the examples you mentioned imo.
Naming two successful technologies says nothing. Your steps don't follow inevitably from one to another and there's a vast universe of failed technologies and stillborn revolutions that prove it. If you want to persuade anyone that vibe coding will ultimately replace traditional coding you need to give reasons you think it belongs with digital cameras and streaming and not with lisp machines, ekranoplans and every attempt at "no code" in history. Otherwise you're just making noise.
EDIT: digital cameras didnt work, until it did. streaming didnt work, until it did. ecommerce didnt work, until it did.