Hi JTB, yup the problem there is with the hole at the very bottom when looking at the top view.

For the weird result at 0.4 radius, you can see that the fillet for the outside edge is overlapping just a little bit with the fillet for the inside of that bottom hole.

Once things are overlapping like that, stuff for determining what is on the "inside" and "outside" of intersection results can tend to get confused.

- Michael