But I'm confused by one thing: is UE 3 tricky when it comes to gameplay? I don't know much about engines, but I'd like to find things out.
Okay, let's start with some basics:
An engine is like the ground work for a game: it's basically the paper on which you write your story (meaning the game as a whole structure of different functions and interactions between objects). Therefore, the engine must contain some standard tools for creating a game, like a graphics renderer (not sure whether that's an actual english word) or the base structure of an AI.
So, by using the tools provided in the engine, you can start writing your own events, graphics, sounds (limited in quality and quantity by the engine) for a game. Movement input, cues for gameplay-pauses like short rendered clips, etc. are all defined by your engine.
So based on what I know about Unreal Engine 3:
We've got the processing capabilities to display a variety of moving objects in gameplay, without eating too much of our graphic units or RAM (example: Arkham City. Combat works smoothly, even though you're sometimes fighting as much as ten armed dudes). An advanced AI-system and renderes environments with dynamic lighting (DmC-screens, GoW).
From what I can tell by the trailers (and that isn't really much) UE3 doesn't need as much loading screens inbetween (in comparison: DMC4 had loading screens everywhere (Warp doors, cutscenes, you know what I am talking about).).
To do all this, they have to cap the framerate at 30fps. MT Framework could've performed without loading by capping at 30 fps, too, but Capcom decided against it, I assume.
Note: this is just my own, very basic understanding on how engines work. Not everything might be true, since I got it explained by an informatics student at my university a few months ago, but to me it makes sense.