Reading through the answered questions actually was pretty informative, and I appreciate that they did it (better if they bring it here!). Basically it sounds like their imperatives were (1) offer a revenue option, as is now industry-standard, (2) preserve some form of mileage option, and (3) stay within whatever their constraint was on value delivered through the program (I assume the new earn structure saves a non-trivial amount, but at the very least they're not adding more value and therefore couldn't just offer revenue as an option while retaining the prior accrual structure as another option). I still think the result is unwieldy enough, and just plain goofy enough, that it's unlikely to stick long term, but it's workable enough to get on with figuring out how to maximize it.