If the utility can be arbitrarily large depending on N, then an arbitrarily large finite skepticism discount can be overcome by considering a sufficiently large N.

Of course a skepticism discount factor that scales with N might be enough to obviate Pascal's Wager.