It's good, but somewhat inaccurate.  In the first section about filtering on member_number, it says that the leaf level of the non-clusterd index has the row ID.  This is true only if the table is a heap table (has no clusterd index on it).  If the table does have a clustered index, the leaf level of the non-clustered index on member_nember constains the clustered key value, which is then used to do a clustered index seek to get to the actual data.