True but anyone (even former developers at Looking Glass) citing Ultima Underworld has a lot to live up too. We've already seen Thief die…
Thief dying isn't really about Looking Glass.
As for expectations, I try to stay realistic. I know I'll never get what I want, because I'm extremely greedy and my own concepts would require a ridiculous amount of time and resources. Also, it would have to align with my own idea of what I thought was great about Underworld. That part does seem to be covered based on the recent interview, though.
I never found the traditional RPG mechanics terribly vital to UU (stats and sheet doesn't really fit with a representation of physical reality in my mind) - and System Shock moved away from it almost entirely.
I did like System Shock 2 in terms of the mechanics, but I think it would have been better served with a less traditional approach. I think the UI and stats implementation sort of worked against the immersion-based gameplay.
Mind, these guys probably know the fallout that would result from a badly done RPG although I have no doubts such veterans of the industry would produce a bad game. I just hope it's not something that's too dumbed down. One of the core aspects of the Underworld games (and its spiritual successor Arx Fatalis) was the complexity and interaction with the environments.
I don't think worrying about a fallout is a smart way to develop your game.
I think if you have a vision you believe in, you should stay the course and pour all the heart and talent into it you can muster.
Fuck the audience, basically.
While such a thing would scare a publisher to death - it's surprisingly effective when you're trying to do art. In fact, if more people worried about their vision and trusted in themselves instead of trying to prevent a poor critical reception, I'd claim we'd be seeing more high quality games and more crappy games overall - but I'd rather want that, than a ton of mediocre and safe games.