With a swept wing you need to consider the geometry of the whole wing not just the root. Otherwise you'd be putting the cg in the same place for a straight wing as for, say, a 60 degree swept wing, and without even doing any analysis that seems pretty self evidently wrong.
The 33% or 1/3 rule of thumb you are using depends on the relative size/effectiveness of the tailplane - something engineers call tail volume. (Its the area of the tail times the distance between the tail and the wing, divided by the wing area and by the mac of the wing). assuming you have a 'reasonably' sized tail relative to the wing, 33% mac sounds a decent first guess.
If the wing is constant sweep and taper, then you'll get a rough idea of the MAC at the 40% span location - the local chord there is about the mac. (Someone's going to correct me with the exact %age, I'm sure, but since you're in rule of thumb territory it's not critical i expect)
Whether the flow is laminar or turbulent, you still need to account for the wing planform.