Is a pretty big claim your title is making.

Your premise about organisational design looking like a game of roulette is pretty weak analogy considering the later is a competitive pursuit against chance. The former being an optimization of communication effectiveness.

Why not state your sources of your model: https://teamtopologies.com?

You have some interesting points about tenancy and domains, but this model may only be effective in large enterprise software orgs (I am assuming from your context) not applicable to other kinds of software companies.