This study published data from 50,000 DNA tests (F adjusted column - Additional file 1: Dataset 1)
It does note that:
There were interesting exceptions to the correlation of inbreeding and health. The Border terrier, Basenji, Collie, and English setter breeds have high inbreeding but low morbidity. Likewise, the Malinois, Pomeranian and Russian Tsvetnaya Bolonka (Russian Toy) have lower inbreeding and high morbidity. These example breeds are neither brachycephalic nor particularly known for extreme morphologies. In the case of healthy breeds with high inbreeding, it may be possible that these breeds have been purged of deleterious alleles as has happened with inbred mouse strains [46]. In the opposite situation (lower inbreeding and high morbidity), the recorded morbidities could be high allele frequency Mendelian diseases or potentially conditions linked to phenotypes under selection in the breed.
You can also look at the plots in the study and see that while a high degree of inbreeding is associated with considerably higher mortality, there's a huge amount of overlap. You can do dog eugenics and get very unhealthy dogs, but you can't do dog eugenics and get dogs that will never have any health conditions (except the worst physiological ones that only exist because of ridiculous amounts of inbreeding).
If you have a general idea of the temperament of the parents, you'll have a better time. The issue is that a lot of a dog's 'personality' is based on genetics, and while you can train it out of them if you start when they are puppies, but it can only do so much. Eg. If you breed an anxious or nervous breed of dog with an energetic and friendly breed dog, there is a good possibility that the puppies will have huge attachment anxiety and will rip up your house if you don't have another dog around, something you could train them out of, but you would literally be fighting their genetics to do so.
Most dog breed owners prefer to train with the genetics of their dog, which can be easier with pure breeds. I say can be because there are a lot of weird things that can and do go on there. Long story short, mutts are usually fine if you have an idea of who the parents were and their temperaments, and can have time to train them. Otherwise it is a completely mixed bag.
Overall mutts have better tendencies towards physical health, but again, it can be a mixed bag.
So you can see here that border collies on average take instruction much better than the average dog, but it's very difficult to determine whether the bordie collie you get will be any good at taking instructions.
They also made a website with the data https://darwinsark.org/muttomics_viz_dashboard/
deleted by creator
They can be, it depends on their ancestry
This study published data from 50,000 DNA tests (F adjusted column - Additional file 1: Dataset 1)
It does note that:
There were interesting exceptions to the correlation of inbreeding and health. The Border terrier, Basenji, Collie, and English setter breeds have high inbreeding but low morbidity. Likewise, the Malinois, Pomeranian and Russian Tsvetnaya Bolonka (Russian Toy) have lower inbreeding and high morbidity. These example breeds are neither brachycephalic nor particularly known for extreme morphologies. In the case of healthy breeds with high inbreeding, it may be possible that these breeds have been purged of deleterious alleles as has happened with inbred mouse strains [46]. In the opposite situation (lower inbreeding and high morbidity), the recorded morbidities could be high allele frequency Mendelian diseases or potentially conditions linked to phenotypes under selection in the breed.
You can also look at the plots in the study and see that while a high degree of inbreeding is associated with considerably higher mortality, there's a huge amount of overlap. You can do dog eugenics and get very unhealthy dogs, but you can't do dog eugenics and get dogs that will never have any health conditions (except the worst physiological ones that only exist because of ridiculous amounts of inbreeding).
If you have a general idea of the temperament of the parents, you'll have a better time. The issue is that a lot of a dog's 'personality' is based on genetics, and while you can train it out of them if you start when they are puppies, but it can only do so much. Eg. If you breed an anxious or nervous breed of dog with an energetic and friendly breed dog, there is a good possibility that the puppies will have huge attachment anxiety and will rip up your house if you don't have another dog around, something you could train them out of, but you would literally be fighting their genetics to do so.
Most dog breed owners prefer to train with the genetics of their dog, which can be easier with pure breeds. I say can be because there are a lot of weird things that can and do go on there. Long story short, mutts are usually fine if you have an idea of who the parents were and their temperaments, and can have time to train them. Otherwise it is a completely mixed bag.
Overall mutts have better tendencies towards physical health, but again, it can be a mixed bag.
The degree to which genetics influence behaviour at the individual dog level is very much overhyped - this is some decent research
https://www.science.org/doi/10.1126/science.abk0639
So you can see here that border collies on average take instruction much better than the average dog, but it's very difficult to determine whether the bordie collie you get will be any good at taking instructions.
They also made a website with the data https://darwinsark.org/muttomics_viz_dashboard/