Yes, definitely fair. Ideally all the models would be trained against "Pi in exactly the configuration I personally prefer", but alas :)