Luckily the wild types of most of the species that were domesticated are still extant. But imagine the wolf was extinct because of human persecution. Conservation could try to substitute their ecologic niche with other wild canines, or create an effigy that can be released into nature in order to redevelop surviving strategies and social behaviour patterns typical for that species, Canis lupus. Interestingly, there was some kind of “breeding-back” effort for the wolf, the Tamaskan dog. It is a crossbreed of less-derived dog breeds with the target to resemble the wolf in appearance but without crossing-in any true wolves. And just like cattle and horses, domestic dogs were domesticated in the middle east but likely experienced introgression by European wolves, making the situation of the Tamaskan very similar to that of other breeding-back attempts. The fact that living wolves still exist makes it possible to see how successful classical "breeding-back" can be.
The breeding was started with a number of sled dogs that were crossed with Siberian Huskies, Alaskan Malmute and German Shepherds in the 1980s. The similarity that has been achieved in some individuals is indeed impressing. Tamaskan dogs have size and proportions very reminiscent of some wolf populations, and also the colouring is very similar to their wild type, with some subtle differences. Regardless of that, Tamaskans still have a number of domestic features. For example, a fifth toe on the hindquarters appears in some individuals, and the face of the Tamaskan differs from that of wolves in being more paedomorph by having a broader snout, larger eyes and a more domed front head, and I am pretty sure the brain volume is reduced as well. These are typical domestication features evident in virtually all dog breeds, just like enlarged udders or dewlaps, a reduced brain volume and snout length are found in most domestic cattle breeds.
Without doubt the Tamaskan is a domestic animal and not a wolf, although it does resemble its wild ancestor. It would be interesting to see a genetic comparison of several dog breeds (including the Tamaskan) with the several wolves from different subspecies, to see if there is any significant difference between the relationship of the dog breeds to the wolf. However, I do not expect the Tamaksan to be closer to the wolf than its founding breeds, for the simple reason that it was selected for a few key genes for phenotypic features, but the whole genome of a wild type consists of more than that.
It would be interesting to see how a population of Tamaskan dogs would do living in the European or North American wilderness. There are a lot of examples of dogs surviving in the wild, so I think the Tamaskan would do so as well. They’d probably evolve hunting strategies and complex social behaviour just like, for example, Dingoes. One could start a dedomestication experiment releasing a population of Tamaskan dogs in a large area with deer and other prey animals and watch evolution doing its work. After some amount of time, we could compare these wild Tamaskans to true wolves. (I don't know if such a project is legally possible, I think it isn't).
What would be the point of that? First of all, such a project itself would be interesting enough. Furthermore, the history of the Tamaskan is surprisingly similar to primitive cattle and horse breeds designed to be aurochs and tarpan proxies. It shows that no matter how authentic an effigy breeding result will get, it will always be different from the genuine wild type. But if man can achieve both “bovine Tamaskans” and “equine Tamaskans” that resemble their wild type just as good as this dog breed does and survive in nature in a similar manner to their wild types, it would be awesome. And several projects are trying to do so. Such cattle and horses would be a prime basis for dedomestication, for letting nature refine them and develop surviving strategies and social behaviour necessary for survival so that they can re-occupy the vacant niche of their extinct ancestors.