What it boils down to is, either the currently accepted c is wrong, by virtue of not being constant, or the data samples are inaccurate due to a lack of sensitivity or bias. Because the data sets do show variations in the speed of light. So, either they are wrong, or they are wrong. If they have calculated the wrong value of c, then there are many other things dependent on c that are wrong. If c actually IS variable, then there are many other things dependent on c that are incorrect. If they want to claim that c is accurate despite the data, then they need a way to explain the variations and abnormalities seen in c.