All that does is show BOTH dog breeds and wolves as tips of the canid tree.
Genetically, the wolf is the same as a dog. They can interbreed, which shows them as the same 'species', and they are similar both morphologically and genetically. There is more diversity among canids than most other haplogroups, yet they have not become reproductively isolated.
The condition of domestication is a man bred trait.. genetically, the wolf and dog are of the same haplogroup, or genotype. IOW, the wolf (many varieties) are just canids, like dogs. They descended from the same ancestor, and became morphologically homogeneous through man made or natural selection.
There is still a lot of variability within canidae, and new breeds (of all canids) continually present themselves. But the SOURCE of that genetic information is deep within the gene pool, and has limited presentation. There is no mechanism to 'create' new traits. Only by 'selecting' the traits, can they become regular occurrences in the new clade or breed.
Again, I described one common mechanism for producing new traits: duplication of genes along with subsequent mutation.
Now, in such a situation, the duplicate would exist in the parents, but the mutation would only show up in the children.
Would an example of such a situation, where we can pinpoint when the mutation occurred and show it did not exist in that gene line prior to that event, would that be enough for you to accept that new traits can be produced?
Alternatively, what sort of evidence would be acceptable for showing that modern traits in dogs did not exist in the ancestral population? Would we need to do a genetic analysis from ancient wolves and how many such would be required for you to be convinced a trait didn't exist in the ancestral population?
Why is it not enough to show that the dog traits don't exist anywhere in the modern wolf population? Or where we can point to when the trait was first expressed? Would you need a genetic analysis of the parents of the dog where the trait was first expressed?