Will changing AttributeHierarchyOrdered to false impare query performance for a large dim

We have a large degenerate dimension based on the fact table.  The fact, and subsequently the dimension, has about 140M rows now.  The process update for the degenerate dimension is taking anywhere from 3-10 hours and am looking at cutting that down using process add.  Greg Galloway has an awesome performance compare on: http://www.artisconsulting.com/blogs/greggalloway/Lists/Posts/Post.aspx?ID=7 which I'm using as a guide to set this up.

The dimension has only four attributes:
Column         Data Type            Notes
ClaimPk        bigint                   high cardinality, this is the dim key
ClaimID        bigint                   high cardinality approx: 85% of ClaimPk
RxNumber     varchar(9)           high cardinality approx: 60% of ClaimPk
ReversalFlag  int                       low cardinality (0 or 1)

I want to change to a ProcessAdd and change the AttributeHierarchyOrdered for the high cardinality attributes to false, but curious if this will hurt query performance in anyway.  Anyone know if AttributeHierarchyOrdered = False will hurt query performance?

(Side note -- I got around that horrible 4Gb limitation for now with a new trick I can post later)

(Another side note -- You might wonder why the low cardinality ReversalFlag attribute is in the fact table.  It is because it can be changed each day for a given fact record and the client wants to see this change reflected in the reports generated from this cube.  I realize the ProcessAdd will not reflect changes, only update or full will, but there is some tolerance to not update ReversalFlag every day of the week, possibly just doing on the weekends, if we can get the daily processing to go much faster)