It does look like something needs clarifying. I get:
> c22=sqrt(0.0431^2+0.0270^2) = 0.0509
> c22+1.3037 = 1.3546
1.3546 and not the 1.3335pu as expected.
Are you sure there are not some missing parallel impedance not shown on the diagram? It may also be that we are adding the impedances arithmetically and not vectorially - do you have R and X for the 11 kV busbars so that we can check this.
Other than the above, in this instance I would contact the software manufacturer and ask him why the results are not the same. If you do this, please post his reply here for future reference.