This looks like a typical case where a scientist publishes a paper that makes certain specific and limited claims, but public relations people at their organization then put out a press release making popularized and overly sensationalized and overly broad claims.
The actual paper seems to be this: Laurent, https://arxiv.org/abs/1106.1068 . They write down a certain perturbation to be added to the electromagnetic field's Lagrangian that violates Lorentz invariance, and then put limits on the size of that term based on observations. Untangling the tensor notation, I think the term they hypothesize is something roughly of the form $\partial(\textbf{E}\cdot\textbf{B})/\partial t$. This violates parity, and also violates Lorentz invariance because the $t$ coordinate is that of a certain preferred frame of reference. The result of this is that the speed of light differs slightly depending on whether the photon's chirality is right- or left-handed. They don't see any such effect.
As far as I can tell, no LQG theorist ever made any prediction that there should be a violation of all these symmetries at the Planck scale. Laurent, who is an observationalist, simply writes down the lowest-order term they can come up with that preserves gauge invariance and rotational symmetry.
Over the last 10-15 years, there has been a history of people hypothesizing that LQG could be used to make testable predictions about the dispersion of the vacuum. However, the people actually working on LQG, like Rovelli, never made any such predictions. More info here:
Does the discreteness of spacetime in canonical approaches imply good bye to STR?
http://physicsforums.com/showpost.php?p=2650816&postcount=28