In transformer it is said that there is no point in having detailed data in the cube. Is that the case with TM1 as well?
The exact reverse. TM1 consolidates (and calculates) on the fly, from base. You can't (without some serious and normally unnecessary messing around) hold data above base.