I thought this through before suggesting it yesterday, so maybe I can provide answers to some questions about it and save Blizzard some trouble.
It might mean that layering will stay a bit longer instead of there being queueing. I don’t like layering either, but it’s better than sitting in a queue, right?
It won’t cause queues in the future except to the extent that people stick around and continue playing where the queues would have caused them to quit. Surely we prefer for them to stick around.
Very roughly, on Grobbulus, it reduced the queues by 5000 between day 2 and day 3. Day 1 queues were comparable to day 2 queues, so I think most of that 5000 reduction is due to the change. For Grobbulus, this equated to a change from 4 hour queues to a 30 minute queue.
Probably post improvements. My understanding is that Herod previously had 8 hour queues; the roughly 4 hour improvement is about the same as the 3.5 hour improvement seen on Grobbulus.
This probably would not have worked on day 1. Server stability is primarily a function of how many people are in the maximally populated zone. Increasing the cap when everyone was still in the starting zone would probably have caused problems. Increasing the cap now, when people are spread out into several zones due to leveling, is less likely to affect server stability.
It should not change the plan to get down to one layer by phase 2. By that point, players will be spread throughout all the zones.
For reference, my suggestion from yesterday, though it sounds like it was also discussed elsewhere as well: