You didn't reply to earlier points. Namely they may lose money on award nights.
And now you are making up percentages to fit your logic. All I am offering are possible explanations.
I used the percentages I did at random. Use whatever numbers you want, the logic is still sound, i.e. that opening up awards raise the possibility they will cross the magical ADR-line, notwithstanding the possibility they'll lose some revenue bookings.
With respect to the possibility they lose money on award nights: 1) there is no evidence as to what the non-ADR reimbursement rate is for a Waldorf, and whether such sum would not be sufficient to cover variable costs, 2) even if it's insufficient, as has been mentioned, award guests spend $ at the hotel (room service, restaurants, upgrades, etc), 3) more importantly, they clearly would not lose money on bookings passed the magical line; in fact they'd gain considerably, and 4) each award booking that gets closer to magical ADR-line has the potential to be reimbursed at ADR. The carrot seems to present a large enough incentive to disincentivize the (seemingly ill-advised) practice of restricting award nights, and that becomes even truer the more the ratio of the current occupancy skews towards award bookings (at the magical crossover, the hotel transforms a larger group of variable rate reimbursements to ADR-reimbursements).
Thus, even if they *do* lose $ on non-ADR reimbursements (which is at least a debateable proposition), the money is lost would pale in comparison to the potential gained by having a bulk of rooms reimbursed at ADR.