User blog comment:Pteriforever/Obliviation/@comment-1782812-20141209002417

Okay, I adjusted the definition pretty seriously.

Without the "breakout rule", it did indeed reach  f_w^w(n), but with it it only reached f_2w^2(n), meaning after the change it didn't even outgrow gigotion. To remove the "breakout rule" would be a major sacrifice in expressibility and elegance, so I left it in and built some higher notations for it

It's been extended and fixed; in addition, it's neater now, with the unary version nicely diagonalising over BEAF linear arrays ^.^