Longevity, so the game isn't technically outdated by the time it was released. Ease of use, lots of documentation online so developers spend more time working instead of learning. Easy licensing, Unreal 4 is free with a simple royalties system for successful games. They don't need to negotiate or purchase a license for another proprietary engine, or spend time building their own. The only engines I know of that fit the above criteria are Unity 5, Source 2, and Unreal 4. With just layman's knowledge I don't think the former two are good choices for a game that ultimately wants to continue to raise its scope in terms of game size and number of players.
If it were possible I would have developed an own engine for EC, to make sure that the massive warfare actually works. Yes I'm full aware that it takes a lot of time and money, but it pays off in the long run. Too bad BHVR said at the beginning that they want to focus on creating the game and not an engine. We have seen how that worked out...
First of all the game developer MUST know what he is doing in case of technology. Even on theoretical basis so he/she is able to use all so called features in the game engine. In the UE 4 there is good gui which is needed to be known or learnt to get all of features. And there is many technical problematics in which UE 4 shines. Just by the way: inverse kinematics. Very robust and up to date iterative math to get very precise results. And the purpose: you can interact with the character you are playing with the game environment very precisly. It is running on the terrain mainly. And visual feedback for environment collisions with "realistic" movement there. The mentioned lighting effects are, ehm ..., not on the top but still it is much much better then in unity. But to be honest: this game is still simple shooter so devs can afford to have UE4 as the more hefty engine to have in the game. The problems will start when this game is going to be so called "massive" ... in other words mmo. Then the UE4 may be huge problem for the performance. But I must be honest again: I dont expect any kind of mmo mark in this game anytime. It is overpriced heavilly even it looks very good. But who cares?
Too much money. Can you imagine what it may cost to develop the game engine for mmo? It is monstrous ammount of money. Even there is many technology blueprints of many features(for example already mentioned inverse kinematic setups, terrain gradient maps, efficient geometry representations, etc., etc. ...) but still it costs so much effort to push the game engine into useful state. And EC is very demanding in this case. It asks for very precise combat/movement evaluations to represent it efficiently.
They're all early access/made by indie devs. And UT4 runs perfectly fine on finished maps, I used to have 70+ fps on a 560ti on the first finished map on medium-high settings, you're either lying or have the worst toaster ever made.
