I just caught back up with this thread and need to apologize for the delay but work has kept me far from anything computer related.
CJR is giving some very good advice as to what could be happening here.
some time back Tim and myself had a discussion about mechanical and static fitment of a boolit.
and how the designs need to differ from each other.
if the collapse occurs in the mechanical stage and the design doesn't support that, it can lead to results like discussed above.
CJR gives the explanation on how the cog/cop relationship can give you fits when it flip flops mid throat so to speak.