Regrettably, references to discounted Harville go over my head, but I have read the Meadow book (whoever posted that he needs an editor is correct). The focus of the book is not on choosing between pools, but on comparing the ROI (sometimes just win; other times win, place and show) and impact value of a great many "angles." He does include win, place, and show percentages in many cases. In the case of favorites going off at 1-1, those percentages are 41%, 64%, and 78%, so the overall probability of finishing exactly second is 23%.

In my mind, whether betting the favorite underneath make sense depends on why you think he might lose. If you think someone else might jump up, it makes more sense than if you think the favorite may not run his race. In the race you are talking about, I was listening but not betting. It seemed like the risk was the latter, that the favorite might not handle the 9f (given that he had never run more than 6.5f and seemed on edge in the paddock and warm-up).