Somebody may make a breakthrough that gets improvement going again, but there's nothing in the fab technology pipeline that looks to pay off big within five years. If there was, it would have to be in prototype now, because the lead times are that long.
So, now what? More emphasis on optimization? Rejection of bloatware?