Yes and no.
If the base is absolutelly level, and the head is absolutelly level, then you are right.
But if one is slightly slanted to one side (beyond the meter accuracy) and the other one is slightly to the other side, you have two times the error and don't know about it.
Using mirrors or a laser, you are adjusting one relative to the other - there will not be addictive errors.
Besides, as Gebhardt pointed, there's lens plane adjustment to put one more variable in the process.
Now, how significant are these errors? I don't know.