I propose that hybrids have a much bigger impact on future keepers than naturally occurring ball python mutations do.
If you don't like the occasional side effects of the spider mutation they are pretty easy to avoid, don't buy or breed a spider. Same goes for any of the dominant/co-dominant mutations. Recessive like caramel is a grayer area. There might come a day when people accidentally produce kinked caramels from unknown hets. People who produce possible het caramels (and I'm one of them) have some impact on future ball python breeders.
But to me hybrids are the caramel case 10 fold. It's only a matter of a very little time before the decedents of hybrids are mixed into the general populations.
Some breeders do a good job of tracking linage and passing that information on but I believe most are like me and often don't take the time. Someone recently posted a pic of two pastel clowns. Was this the first year for that combo or was last year? At any rate it's a very rare and new morph and the current owner didn't know if the two animals are siblings or not. Not really a big deal or a bad reflection on either the seller or the purchaser but just an example of how quickly linage information is lost even on very special animals. If there aren't already unknown hybrids in the general ball python population it's only a mater of a few more years until there are.
On another forum someone who I believe has a scientific background (I don't) pointed out some studies where crosses of related but different looking plants or animals are made and then those crosses are bred together and the ratio of offspring that look like each of the two grandparents tells how many mutations separate the original plants or animals. The point was that a surprisingly small number of mutations (5 or 6) might separate the appearance of two species but I'm not convinced those visible mutations are the only differences. While we might get some animals that could pass for pure balls or pure burms fairly quickly from breeding hybrids together or even quicker by breeding them back to either parent species I believe there would still be many harder to see genetic differences between those hybrid descendants and the original species. After probably millions of years of separation these other differences could well effect chemicals needed for breeding or cycling or incompatibilities causing health issues or reduced life span.
The cost of us enjoying a very few problematic naturally occurring recessive ball python mutations that future ball breeders might not want to reproduce (caramel is the only one I can think of, and I for one would still be very happy to produce a caramel) is that some future ball python breeders might want to track lineages for caramel occurrence. However, the cost of the first ball python hybrid is that all future ball python breeders may have many difficult to detect incompatible genes to worry about that might shorten the life or reduce the breeding effectiveness of their animals and at the very least change a few of the aspects of what makes each species unique.
