But it doesn't, is the thing. Taking on an extra curse only gives more benefit If your odds of survival are not negatively impacted thereby. Tactically, it's a bad move to take more curses without spending the Remittance Value you're allotted for them. If Hunger had not picked up the Ring Hunger, he wouldn't've been able to take, say, Blood Halo, which gave +5 Accursed Favor, 3 more than the runner-up option. If Accursed Favor is benefit to the Accursed, you're stating that the choice of not taking the Ring- granting +1 Accursed Favor, and subsequently providing at least an additional -3 Accursed Favor relative to the alternative- would've benefited the Accursed, despite, on net, having lower Accursed Favor.
In order to make 'Accursed Favor Measures Benefit' consistent, we have to conclude it measures immediate benefit, without accounting for future actions we'll eventually take, or at least without accounting for anything specific or particularly unlikely- since otherwise, the +Accursed Favor at character creation doesn't make sense. If so, then clearly the Accursed Favor value of Vengeance isn't accounting for the long-term benefits of the Forebearer removing the doom of Tyranny, since those are unlikely to manifest (99% chance of death first) and specific (what if he fights the accursed instead!1!). This is supported by how Vengeance isn't at +several billion Accursed Favor, since the benefit to the Accursed of completely removing a Curse is clearly transfinitely superior to the benefit of taking on a transfinitely-tiny spec, and by how Vengeance isn't at +several hundred Accursed Favor, since Hunger'd probably pick up several other Accursed-Favor-Worthy methods of mitigation in his long travels.
At most, then, the 'Accursed Favor = Benefit' side can argue that Hunger choosing Freedom gives the Accursed immediate benefit roughly equivalent to our taking Fisher King, whereas Vengeance gives the Accursed immediate benefit... consisting of our taking Fisher King. (plus change.)