Deal with January 2017 false positive scenarios
Now that we are close to 3.0 and
feature/stretch being merged, let's down-prioritize new robustness issues in
devel in the sense that we ignore the issue unless we find proof that it also affects
feature/stretch. This might mean that we ignore an issue that actually affects
feature/stretch but isn't visible due to something breaking earlier in the affected tests, but let's prefer that risk over identifying and tracking issues that don't affect Tails after Jessie.