It's not always true that 'professional' amphibian ex-situ efforts assess the genetic heritage of animals used in projects. The vast majority use animals taken from rapidly disappearing populations, so although they are all from the same population, all animals are rarely genotyped etc. For most projects, the effective population size is absolutely tiny, too, with very few founder animals and many closely related offspring. These projects rarely if ever subject animals to natural selection, other than culling excess or obviously ill animals, often because zoos are uncomfortable doing this and/or do not have the licenses to subject animals to predation etc, or because so few offspring are generated they can't afford to lose any. Despite all this, reintroduction projects have been successful or semi-successful in the past, and look at the number of naturalised populations of alien species that started from small populations of dumped pets, often CB (e.g. banded newts in Spain, alpine newts in the UK etc).
Despite this, I do agree with Mark that any pet animals are no longer part of wild populations and that it is not ethical to attempt reintroductions in one's back garden, for both ecological/genetic reasons (disease, non-native genes, altering fitness of surviving wild populations by introduction of new genes etc) and welfare (many animals will die, particularly as there is a good reason why native populations have disappeared). Exceptions may exist where extremely rare species only exist in captive populations (imagine if the golden toad was being bred by a hobbyist).
I think that most CB hobbyist populations, particularly when large numbers of animals are reared in relatively natural tanks and where juveniles are allowed to 'get on with it', rather than being hand-fed, populations can remain relatively fit. This is not to say that they are entirely the same as wild animals, but they are perfectly viable. As long as populations are managed to an extent (stud-books, keeping track of which animals are related to which and how, avoidance of line-breeding), they can remain relatively similar to wild animals. In fact, the altered selection pressures in captivity are likely to produce animals that are much better at coping in aquaria etc and make healthier, happier pets.
Many of the differences we see between wild and CB animals are highly unlikely to be genetic (there haven't been enough generations to see these effects), and are much more likely to be due to plasticity in the salamanders in response to different environmental conditions. Most species develop and grow much faster in captivity due to more plentiful food and more clement conditions, and are fed diets that are vastly less varied than wild animals receive. The lack of winter-summer cycling has important effects on morphology and fecundity, too. The list goes on.
Therefore, I think the main reasons to keep genetically healthy populations are aesthetic (if we like the look of the them as wild-type) and sustainability-related (ideally, we want to be independent of new imports in case they stop and so we are not responsible for declines and extinctions), but providing appropriate conditions long-term will have an effect on the viability of captive animals long before responses to selection kick in. Mallorcan midwife toads (Alytes muletensis) have been shown to lose anti-predator behaviour after around 8 generations, but few captive salamander populations are past F1 or F2 and even longer would be needed to start seeing more obvious genetic problems that could threaten viability.
This brings us back to hybridisation. Reasons to avoid this are either aesthetic (we prefer the idea/look of having wild-type animals, rather than hodge-podges) or sustainability-related (hybrids are often less fertile, massively more fertile, sterile or generate sterile F2s), all of which could seriously damage the long-term viability of captive populations, particularly as most species are capable of living a LONG time and of generating MANY offspring. If people want to generate hybrids on a large scale, I won't be buying them, but as long as they are advertised honestly and don't get interbred with pure species newts (unlikely, in reality) I don't think it is particularly damaging.
C
PS, the pictured animals look very healthy. Out of interest, were they easier to rear than non-hybrids?