This is not established fact, but an asserted belief. What evidence do you have that the genes that were revealed as 'new' traits were NOT already there? Selection acts upon EXISTING variability, which means it has to be there to 'select'.
And that is an unknown assumption. The logical, biological FACT, is that the variability we see, now, was ALREADY PRESENT, in the ancestral canid.
Asserting a belief in 'new created genes!', with no evidence, does not compel a conclusion of common descent.
Not so. The only 'new breeds!' we see, from the tips of phylogenetic trees (like chihuahuas and great danes) are cross bred with other breeds, to inject some diversity that is not there, in the low diversity breed.
You can breed purebred chihuahuas for multiple generations, and you will not get great dane traits. Only if you could cross breed a chihuahua and great dane, can you inject their respective traits into the child populations. (That is a difficult mental image!
..)
So you assume and assert. But the facts say otherwise. There was not enough time for 'mutation+time!' to "create" all the diversity in domestic dogs. That variability could only have been in place in the ancestral canid parent.
asserted and believed, with no evidence.
Traits that come up that were not seen before, is the Big Question. Were they already there, in the ancestor, waiting to come up in the lottery pull of genetics? Or, were these traits 'created' by some undefined, unobserved mechanism, usually shrouded in 'time + mutation!?'
Evidence for either 'theory' must be presented, not merely assertions of belief.
All the evidence, in observable, repeatable science, says the traits/genes were there, in the ancestral population. There is no evidence that any of the canid clades were formed by mutation. That 'theory', or more accurately, 'belief', would have to have hard evidence to support it, not just plausibility or conjecture.