I think the author had that insight, just chose not to use that route because it wasn't necessary. For either a human or machine solver, the trial and error ro

I think the author had that insight, just chose not to use that route because it wasn't necessary. For either a human or machine solver, the trial and error route may well be simpler / less code / faster at runtime than implementing and executing every way to recognize a set of N peers for N values.