It is very common and very useful in hatcheries to perform break outs on the eggs that didnt hatch. In that way we can identify if the eggs were fertile to start with, and when the embryo died and if anything unusual caused the mortality.
We usually calculate the number of for instance early deads into a percentage based on the eggs set. So if we have 150 eggs on a tray and we find 3 early deads, we calculate (3/150)*100 = 2% early deads. And based on the age of the flock, the storage conditions etc we then decide if this level of early deads is acceptable or if something strange is going on.
However, we have to realize that relating the level of in this case early deads to the number of eggs set can be misleading. As infertile eggs cannot contain an embryo and therefore the embryo cannot die, relating the number of early deads to total eggs set can give us a wrong perception of the reality.
If we have 100% fertility, than with 3 early deads it means that from the 150 live embryos 3 of them died, being 2%. But if the fertility is only 50%, only 75 of the eggs contain an embryo that can die, and if 3 of those embryos die the real level of early deads is not 2% but 4%, (3/75)*100. And perhaps we do not consider it a problem if 2% of our embryos are dying at an early stage, but that might be different if we realize that actually 4% of them died.
Of course this holds for the other characteristics as middle and late deads, malformations, malpositions etc as well. Infertile eggs do not contain an embryo that can die, so we should relate the abnormalities that we find in unhatched eggs to the eggs that could die, the fertile ones. And not to the total eggs as that can be misleading.