Those comparisons are from 1 point in time, 120 hours out. A lot can go wrong with a forecast 5 days out with any or all of them. I "think" I recall seeing a comparison less than 120 hours out and there was a marked unification of all the model forecasts as would be expected.
Personally, with all the things that can and do go wrong with forecasts, I think this is one of those things (best model) that rests with anecdotal evidence as flawed as that may often be. I mean we all can probably agree that the Euro used to be the best model based on our experience, but can likewise all agree now that it is not as good as it was based on experience. Jmho