I have great respect for what you, and every other serious breeder does. As a breeder myself, I think the agriculture industry in general puts too little focus on genetics, and too much on non-structural band-aid solutions like pesticides.
If I understand correctly, though, your 500,000 crosses are the result of 1 generation of crossing between all of your hop cultivars with one of two Corvalis males? If this is the case, the low yield of dwarves would be understandable, and presumably associated with mutations, unless air-borne pollen from a wild dwaf male contaminated your lot. Not sure if you used open pollination or not.
This strategy offers the advantage of having more time and resources to determine the very best from a very large sample pool. It's disadvantage, however, is the low yield of novel traits. Maybe we could get a new hop cultivar that yields twice as much as the leading one, just by combining alleles from X, Y, and Z cultivars. But if these alleles are absent from the males, single-generation crosses would never be able to have all three of these, nor even two. While even if you increased to 3 males and each had X, Y, or Z, single-generation crosses would still never be able to unlock their full potential, because you'd never get all three in the same specimen. Even if you did millions of crosses. Many traits in nature are additive, so it's very possible that some desirable traits in hops are as well.
But it seems like there is very little literature on the genetics of hops. I guess amateurs tend to lack the theory behind it all, and professionals tend to keep the details to themselves. As with many other species I grow, I guess I'll have to do my own trials to learn more.