It's a great idea. I love it. However, there are only two ways I can see something like that getting built: the large way or the small way. The large way would be as a tool designed to work with one of the major languages/environments. It would have to deal with all the bloatedness that implies, plus the fact that those environments were never designed to provide this sort of transparency or hackability. It would probably be a large corporate project that took years and cost millions; this probably won't happen. The small way would be because one or more hackers decided to build something new that stayed close to the hardware and was designed for this sort of thing. That's much more doable, but puts you well out of the mainstream for a long time.

I wouldn't be surprised if people eventually come around to this point of view. It will take a while, though. The current thought is that parallelism will save everything. It could take a decade before people figure out that that isn't going to happen. Maybe then there will be a renaissance of going back and rebuilding things from the ground up. But I'm rambling.

"People who are serious about software have to be serious about hardware"... that Alan Kay has a gift for putting his finger on things.