I haven't fully understood why you are looking for something different than dimensionality() ('..but for pivot table totals rows). I always assumed dimensionality() was designed for pivot tables total rows.
dimensionality() should give you the level of your partial sums, if you need to distinguish between lines of the same level, you can check on dimension values.
Using NoOfRows() doesn't seem very robust to me (when adding / removing some Title2 values).
You can probably work out a solution using a pick() on dimensionality()+1, then maybe using another pick/match for the single expressions in each level based on dimension value.