I wonder if the same study was carried out in say twenty or thirty years time, whether there would be such apparent differences between show/working dogs, particularly if the breeding of working strain dogs for the pet market by breeders motivated by money (ie not genuine breeders who work their dogs) continues and has the same detrimental effect as the commercial breeding of show type dogs.
I take slight issue with the apparent assumption that any breeder of working strain dogs who doesn't work their dogs is motivated by money.
However the truth of the matter is that any breeding that does not have good temperament as its top, or nearly top, priority is likely to cause issues - this could just as easily be working breeder chasing the ultimate high drive working dog, as the issue of past show champions being of uncertain temperament but otherwise of perfect physical conformity.
The problems are magnified by puppy farms and of course the huge numbers of pups that one stud dog can sire.
It is not altogether supprising that Mark found links between certain colours and temperamental issues - if you threw enough resources at a study you might even be able to trace the individual dogs that lead to the issue. The priority though is to look forward and encourage breeding from only temperamentally sound dogs/bitches - for both strains of the breed.