[Shlomoshun]

Here's my two cents. First, I'm assuming populations are mostly blanced, like at 40%/30%/30% for the three sides. If it's worse than that, they'll need PS1 type faction rewards to get the populations in balance. Assuming they are at 40/30/30 at worst, the 40% one shouldn't be able to dominate any of the weaker factions unless the 2nd faction is also helping them out. So balancing out the faction dominance should be built around encouraging #1 to focus on #2 and #2 to focus on #1, while allowing #3 to build itself up. As long as they incentivize rewards this way, none of them should be able to really ever pull away from the other 2 by that much...

Personally, I like the idea of carrots for the underdogs more than i like the idea of Sticks for the stronger side...but the right answer should probably involve both to some extent.
I dislike resource diminshing returns, since the idea of the game is to get control and thus resources. Getting less for holding more land seems counterintuitive.

Combination Carrot/Stick methods:
One way they can provide some balance is by smart placement of resources. The middle ground should provide more what i will term 'luxury' type resources. Things that fund 'extra' types of efforts, not those that fund the basics of your war making capability. The hexes close to each sanctuary should provide more basic resources that do most of the fueling of the war efforts, that way you don't really cut down on any factions ability to fight by much until they are significantly pushed back. Topographically, areas near the bases should be designed to help the weaker faction take them back....Whether thats through providing protection for infantry flooding in from the sanctuary, back doors to bases that face towards the sanctuary so that faction has the more direct route in, etc..., the goal should be to allow the weaker side to have the ability to take back those areas more easily with the more simple vehicles and tactics necessary. Another combo carrot/stick proposal would be that although they have mentioned that the more adjacent hexes you control, the quicker the hacking, what if proximity to one factions sanctuary also had an effect on their hacking speed. That would mean taking a hex right next to another factions sanctuary would take a long time of control, while them taking it back would happen quite quickly for that faction....2 hexes away from the sanctuary would be less so and so forth... This makes sense from the view that areas near a sanctuary would be more loyal to that faction, and thus more easily controllable by them...

Carrot/Reward methods of faction balancing:
Though these would help the losing faction, they would also apply to the the middle sized faction to focus on the winner and not so much the loser.
One great subtle suggestion that I liked from earlier was the direction of auto-generated missions slanting away from the weakest faction, so the auto-missions for the strongest faction would push them more towards #2, and #2 would be pushed towards #1, thus allowing #3 to recover it's footing. Other possible carrots could be more direct, such as XP/resource increases for fighting near your sanctuary, or just for fighting the #1 faction (for both the smallest and the middle factions to encourage them to work mostly towards bringing down #1) or bonus damage when fighting against the winning faction (again for both the smallest and the middle sized faction).

I think if they can just incentivize factions #2 and #3 to mostly aim towards #1 at all times, you'll see that domination is difficult to obtain unless the server has a huge pop-imbalance, which is sort of a different issue altogether.