Thereās probably many other in-game mechanisms that could be tweeked. Sounds like Blizzard has had this problem for a long time, so it would be worth investing the time into testing these things out.
Without being technically inclined when it comes to servers and big data, have you considered having two primary types of servers: 1) Horde and 2) Alliance? You could make a little more granular with Alliance-PVE or Horde-PVP. You could then think of the possibilities of what is and isnāt possible from there. Handles the point that servers naturally gravitate toward this state over time anyway, and allows the communities that form to remain in static server. The tricky part would be integrating the two servers so players from each faction can integrate into a layer or similar temporary state when appropriate i.e. world PVP. Donāt know if the world merging concept would be viable, just a thought.
You donāt need 50/50. I donāt care about 50/50, I was always fine with 70/30 or even 80/20 as long as I had an intact community. My problem was always that 80/20 turned into 99/1.
It would support it. By adding incentives for āsomeā players to prefer the smaller faction. Scaling that incentive inversely to the imbalance would further pressure things away from collapsing. Currently, all the incentives point away from minority factions, so itās not surprising that things trend this way.
I also think removing paid transfers entirely would help slow server collapse. But its obviously too late for that in TBC.