A little late for that discussion, but I only just got to try out the demo, and my conclusion is that the massive negativity generated around it is probably wrong. Some points:
* Just like the first game, the game is flawed in certain respects. The graphics aren't as bland as in the first game, but it is too demanding for my rather high-end PC - and that is a plain mistake. The design also feels a bit crowded, "blocky" and old fashioned in some ways.
* Although the demo is too short to show it, I can feel the depth the game will have because of given by choices, and the multiple solutions approach and not at least: because of the the very fact that the game not only has visual design, but a social and narrative design - a constructed society, history, political situation etc as a backdrop. That last thing is very important - what other game has it?
* The interface is not too dumbed down. In fact it is still too complex. The people complaining are people with an unusual tolerance for complexity accounting for approximately 2% of gaming population (not hardcore gamers like those who read Bluesnews, that is.)
* People forget that the first game was largly forgotten for a while and then grown larger and larger in the last year or so, due evident lack of innovation in the genre and with the benefit of hindsight as to what has actually happened. The original demo would also disappoint to the same extent today without this history.
* This team can probably only make flawed games: the graphics, the lack of sensitivity to things that would frustrate the user and other things, shows that this is an "academic" approach to the idea of making a game. They don't start with the material of digital objects and see what you can shape them into, like Unreal T and Quake. They start with an abstract idea of what the game should be like, and then try to press the game into that mould. What you see is what you get with that appraoch. A game with many flaws due to this strain, but with one important difference: it is made with a important vision, which something very rare.
* We have a new ideological conflict brewing with regard to FPS games: should it be multilinear and based on "emergent" narratives or should it be (semi) linear with scripted events, like the Half-Life makers say (and which also say that multilinearity is *mistake*). I personally think that Deus Ex way of doing has and will show its success. Philosophically it seems to be more important than the Half-Life approach. The Half-Life people will be/is proven wrong about their assertin by Deus Ex I and II. (I doubt if there are much "emergent" narratives in Deux Ex II though. I doubt very much if the AI is good enough for that.
* I think Deus Ex II will be a good game to play, for the same reasons that the first game was good. Because it will show the validity of this approach, it will also be important to gaming.