cardinality in dynamic cubes

[redmist]

i wanted to check if anyone has any insight into the importance of cardinality in cube designer/dynamic cube

i have a fairly large cube with several snowflaking dimensions. Few of the dimensions have joins with the fact to get the necessary information (changing the DW design is not an option, neither is modifying the cube requirements). Joins in these snowflaking dimensions have all been defined as 1:1 (even the ones with the fact - dimension D and fact F). Since the fact is already been brought in this dimension, a join between the dimesion and measure has not been defined. In the implementation tab of the cube, the cardinality between dimension D and Fact F is 1:1, which seems logical since it would have got this information from the dimension.
Now the cube is functioning as expected, data is as expected, rolling up and aggregating as expected in all dimensions.
Does it not need the cardinality between dimension D and Fact F to be 1:n. appreciate if anyone can provide clarity on this
thanks