I'll give it a shot.

1) Not at this time.

2) Probably not ever.

3) Doesn't seem to be a question.

The problem is not computing power but rather an inability to get the necessary input data. Things like the state of stress for billions of points along the fault in 3-D, the exact nature of the fault surface at all points, the strength of the various strata. Little things like that.

IMHO, of course.

Roger