First of all, I'd like to say anyone should stop using "D&D was never for Verisimilitude". this is really a poor argument, and a very wrong one too. You could say that D&D never trully pushed for deep simulation, wich is something really different, and you can argue that D&D always had it's flaws on the verisimilitude side, I agree.
Also, when talking about simulation, you just can't assert that something is verisimilar because I can easily explain it. I could be able to find very strong end verisimilar points for a chicken to fly high in the sky, but that wont turn chickens into howaks regarding the flying capability.
Now, I'm not saying I think 4E will just suck because of a lack of verisimilitude. Mechanics abstraction and balancing are core to the value of a good rule system, so 4E could possibly end up being a great game even if it lacked simulation at all. But if you want to discuss the degree of simulation around...
Consider we are talking about a lot of potential verisimilutude flaws in 4E. Well, actually we know only a very small portion of the actual 4E ruleset. So there's no point in counting 3.XE's (a game you have been playing for 8 years) vs 4E's number of simulation flaws. 3e would just win for the sake of statistic's laws, at the moment.
Also, designer explicitly said "If something was not fun, we've removed it from the game". Now, in my experience, simulation always needs, layered, hard to remember, case by case rules (grapple enybody?), and usually simulation is not fun at all.
So it is pretty simple. Streamlined rules, more fun, less verisimilitude. I'm not in the mood for lenghty digressions on every rule of the game, sorry if I could look like rude, not my intent
