I'd say it's more a question of forecasted demand compared to advance bookings. For example, I've been looking for tickets to SFO this fall, but the lowest bucket open right now is L. The inference I take from that is that DL is happy with the ABK levels at the moment, and sees no reason to open U and T up for sale (even though I'd qualify for the U and T fare rules this far out). So I'm holding out, in the hope that travel slows down and DL decides to open up the lower buckets.

The buckets tend to be nested as well. That is, if I buy a S-fare ticket, it lowers L, U and T by one seat as well, but not higher buckets. So when you see "only 2 tickets left at this price," it means that the next two tickets at or above that fare class sold will zero out and thus close the bucket.

The other complicating factor is that it's an O&D issue, not just a local issue. So let's say you're flying JAN-ATL-LHR. You'd need to have inventory for both JAN-ATL and ATL-LHR. But when you look at the "value" of a JAN-LHR ticket versus a JAN-ATL ticket, the former will bring in far more network revenue. So you might see JAN-ATL at Y0, while JAN-LHR was at B9. Travel agents - clever little thieves that they are - have tried to game this system by booking JAN-LHR, then cancelling the ATL-LHR segment in order to steal lower inventories for local JAN-ATL segment. In response, airlines "marry" the segments so you can't cancel them independently.

So in the end, there's not a very reliable system for predicting when buckets will open or close. Best I can suggest is that if it's a natural high demand period (Thanksgiving), buy early. But if it's just a normal weekend, you can roll the dice and see if the U and T fares will open up.