Good hint!
The Isopach line should by theory not cross itself and the data prep - loop remove didn't find anything.
But after running the "data prep - filter vertices" on the line it removed several hundred vertices and it did now do the volume computation.
For another test I then exploded the original line again and did run the "data prep - zero length removal" with a filter of 1 mm and it did remove 25 segments. After rejoining the rest of it, it also did do the computation.
So it seems there is a test for small segments within the boundary. But instead of spitting out just Zero it should tell you why it won't compute. After all, it does accept the boundary, it won't accept open line work for instance.
And it does compute the area within the boundary, but maybe it's just taking that information from the boundary property tab.