Helmholtz-anisotropic-vector-mms-p1p1cg-unstructured fails Intel test

[Tim Greaves]

Helmholtz-anisotropic-vector-mms-p1p1cg-unstructured is failing one test on the Intel buildbot:

Helmholtz-anisotropic-vector-mms-p1p1cg-unstructured: Assigning cd_convergence_isocomm = [[0.41967802829621298, 0.25669519095530213]]

    <test name="cd_convergence_isocomm: L2 order > 0.5" language="python">
assert(cd_convergence_isocomm[0][0] &gt; 0.5)
    </test>

Helmholtz-anisotropic-vector-mms-p1p1cg-unstructured: Running cd_convergence_isocomm: L2 order > 0.5:
Helmholtz-anisotropic-vector-mms-p1p1cg-unstructured: failure.

This test was owned by Jonny Bull, who has now left. Could someone else take a look at this please and decide if the tolerances can be loosened for this one?

** Targetted for fixing before 4.1.11 is released