I mean it kind of obviously is if the plane is going to be full. Predict Y will be full, take a few Y pax from the waitlist. Predict PE will be full, take a few of those. Doesn't have to be more than that. I suppose taking money of one Y pax to upgrade to PE and one PE pax to upgrade to J would be optimal. For money upgrades there's a profit and a cost associated with each decision, and avios have a rather fixed conversion rate. I'd be interested in knowing how much an upgrade voucher is worth in this calculus.

Now it gets interesting if lower cabins aren't going to be full. I wonder if there's a fancy algorithm saying "Y pax get 5 points and PE pax get 7 points for travel class". I somehow doubt it. I'd expect it to be € all the way down.