There are several possible complications which might fall under "incompatibility" when trying to combine mutation.
One mutation might cover up the expression of another. For example, how would you know if a leucistic was also a genetic stripe since there shouldn't be any pigment left to show the stripe. I think the term for this is epistasis. It looks like leucistic would be epistatic over most anything but albino because it would completely mask the appearance of most other mutant genes. I think this is what you might have heard about caramel and axanthic. The thinking is probably that axanthic would take away the yellow that caramel seems to put in but I'm not sure that caramel doesn't also reduce black so it might actually be a nice true ghost like combo leaving a faded gray and white snake. You just never know for sure how two genes might interact until someone tires it. Maybe caramel adds a yellow that axanthic can’t remove and caramel will largely cover up axanthic.
Another interaction might be that the combined effect of two genes might be fatal. The mutations are messing with the chemistry of the animal and even if each single mutation doesn't have an adverse effect on health it might turn out that the combined effect could. This might be what they are getting it with what you heard about albino and lavender. I think that RDR has yet to get a hatchling from trying this cross. However, it could well just be bad luck (bad clutches happen) as there can't have been too many attempts yet and I think this last year he used a het on one side so should have had the potential to get some single het babies which should not have died due to incompatible mutant genes.
It is also possible that we will find some genetic surprises once we start trying to combine more morphs. Perhaps we will find different mutations which turn out to be the same gene. For example, it might turn out that albino and caramel are alleles - different mutations of the same gene. If this is the case then when you cross an albino to a caramel you shouldn't get any normals because neither parent has a normal copy of the common gene. I crossed a couple possible hets this year and got all normals but I could well have missed on one or both parents (66% chance het caramel X 50% chance het albino). I've not heard of anyone crossing the homozygous animals but that would be the surest way to check for alleles. Maybe spider and pinstripe are alleles but without a proven homozygous of either type the way to test that would be if the spinner ever produces normals or not (same for pewter and ebony).
It is also possible that two mutations might be different genes that are linked by being close together on the same chromosome. In this scenario you can make the normal looking (if recessive, mutant looking if some type of dominant mutations) double hets just fine but it is difficult to get more than two copies total of the two genes combined into one animal. A crossover is required between the two genes to create a copy of the common chromosome with both mutations. Depending on how close together the mutations are just the right crossover might take a long time working with double hets. And if both mutations are recessive you would be hard pressed to identify the first animal with one crossover short of breeding tests. I forget how many chromosomes pythons have but the number of mutations is getting high enough that we will most likely see linkage eventually.
There is also the possibility that a combination just will not look as nice as we hoped (there have also been some that turned out better than hoped). Some speculate that an albino pied might not have enough contrast to rival a regular pied but until one is produced we just don't know and even then it will be a matter of personal preference.