User blog comment:Boboris02/MBOT/@comment-24920136-20161219043958/@comment-30754445-20161219214452

But that was just a specific example. One doesn't need diagonalization to define f3(x), so it isn't surprising that it doesn't show up.

He also gave an example of Bowers 4-Arrays, which looks fundamentally different. It seems to be following completely different principles. I'm skeptical as to whether this definition actually works, but I also find it hard to believe that Boboris attempted to model an ω²-level function without even diagonalizing once.