Gauss-Kronrod may have a few problems
It looks like the JACOBI_21 scheme in GaussKronrodPts.h is problematic.  The number of points should be odd, but this scheme doesn't have a node at x=0 as it should.
- Put in error trapping for consistency on number of nodes at least.
- Do a sanity check on the other methods as well.
See here for more information.
Edited  by James Sutherland