Manhattan definitely better for assumption family questions. There were also some points that I can't believe LRB never covered, like the range of probability on certain questions, like Must Be True.
Ya, I think Assumption Qs are super important to master because they pop up on multiple question types. Same with Flaw. And what do you mean, "range of probability" and also as it relates to MBT?? I would like to see more of what you think of that...
That's one thing I wanted to make sure of, by reading MLSAT- and BP Shinners wrote of this on his semi-weekly office hours thread- that there are those who get good scores and they can identify the reasoning, the Q Type, the method of attack relative to the type- but that the people who really break through and go 170> "see the matrix," as he called it. They see where the formulas for solving each Q break down, and they see what the LSAT as a whole is trying to do.
I think this comes through a lot of practice and exposure.