Yes, that makes sense to me. --> if you don't specify, then it will use the currently active user frame.
That is what I observed when I test it in Roboguide. Thanks for confirming.
I also noticed that initially, the resulting position was not what I expected (the sum of x,y,z,w,p,r elements). The reason was that the system variable $OFFSET_CART was set to True. Things started working as I expected after changing that value to False.
from Fanuc manual:
If $OFFSET_CART is TRUE, offsets for Cartesian positions are treated as frames and used to
pre-multiply positions. If this is FALSE, offsets for XYZQPR positions are added field by field