I am not aware of any solution provided other than working with the override process which is the alternate iterative calculation method approach.

In all my years with Lacerte, I was surprised this wasn't a corrected calculation.  I also don't prefer to override, but Lacerte's outcome wasn't right.