I've avoided these rhetorical questions but this deserves a PSA, lol.
until they try to map out alleles related to various traits, which we as a species are capable of doing with existing methods and technology, this is all amateur hour. Selective breeding based on what you can see is caveman methodology. There is nothing special or skillful about that. Anyone with common sense and proper resources can be 'good' at that. Anyone can cross breed any number of plants, observe offspring and select the best pairing in hindsight with the best ratio of results.
If you map out the alleles for any trait, there wouldn't be any guesswork and no need for experimenting with multiple combinations of parents etc. You could match compatible genetic code for the outcomes you want and predictably get them or at least get them in a known maximum ratio. Variety of life is not mystical. It's based on a very large, yet finite, number of combinations. It's simply a matter of mathematical probability.
inbreeding even one generation can have negative effects. Genetic depression is a real concern. We need more geneticists with a modern educaton and work experience in this field. All we get now are methods that were available in the 1800s. Monkey see; monkey do level of sophistication is lacking.
The people behind the current offerings probably don't even know how to use a basic punnett square. When this sub-market gets modernized, great things will happen by comparison.