Interesting study, though the study didn't prove that recessive color is the direct cause of the problems. Quote “Because chocolate color is recessive in dogs, the gene for this color must be present in both parents for their puppies to be chocolate. Breeders targeting this color may, therefore, be more likely to breed only Labradors carrying the chocolate coat gene. It may be that the resulting reduced gene pool includes a higher proportion of genes conducive to ear and skin conditions.”
It would be interesting however, if they can track recessive colored puppies and normal colored puppies from the same litters (i.e. control the parents of the puppies, select only puppies from parents that produce both colors for the study). That way, it will be like the study they did with twin babies to determine how much intelligence is environmental vs hereditary. However even this way, it would be hard to get fair data since there may be a difference in owner groups --- i.e. those that choose recessive colors vs normal colors may have a statistically significant difference in the ways they raise and care for dogs, and thus resulting in different chances of those dogs getting problems down the road that is not due to colors.