If you are trying to determine the area of a circle to within 2 siginificant digits and your measurement might be off by up to 10%, it is pointless to calculate pi to 200 digits. :shrugs:
I stopped tracking results back in August of last year. The only result I care to look for now is one that contradicts ultramel. No takers yet.
We are now at the point where I believe misidentification, sperm retention, or a spontaneous mutation are each much more likely to cause a "false result" than anything else. That is, I believe you are as likely to get normals from an amel to amel breeding as you are to get normals from an ultra to amel breeding.
I don't think that lack of disproof in itself is proof. However, if ultramel were inaccurate, there should be inconsistent results coming from every direction. I have yet to hear a single one.
Without actual sequencing you can't "prove" anything. But who cares? We are breeding these animals to acheive certain genotypes and want to be able to "predict" what we'll get.
Nobody has ever proven that amel is a simple recessive... it could be two linked genes. But the fact is that even if it is, how does it totally screw up your breeding plans when you rely on a simple Punnett square to predict the results? It doesn't. Other screwups, or just plain bad luck, are much more likely to cause your results to be different than the prediction.
Same goes for every other known trait. At some point you just have to decide to go with what is a practical and workable description and just go with it until/unless something better can displace it.
The alternative is to pretend for the next 6-10 years that we don't know, and just breed stuff randomly and act surprised at the results. Who is actually going to waste their time trying to "prove" ultramel? For what? To what end? What is to be gained by being "really really superduper even more sure that they really are alleles and not just something that acts like alleles all the time?"