Do you have actual examples of this?

Your solution is to basically do a re-write when scale becomes a problem. Which is the textbook example of something that sounds good but never works

On the other hand I can't think of a business that failed b/c it failed to scale :)

Because you never read about their failures like that. For users they might simply become not interesting, because they fail to deliver good features reliably, because, unbeknownst to the users, they are busy working on their scalable architecture all day.

The final statement rarely is that they over-engineered it and this failed to build an interesting service.