At least in terms of a back-of-the-envelope calculation, what is a believable range for the number of gene changes between parent and child species to represent speciation between species of megafauna?
To paraphrase an old saying, gene changes should be weighed not counted. There are a plethora of types of gene changes, and they vary by orders of magnitude in their impact. In particular, which genes are affected has a huge impact. Some point mutations (change in a single or a few nucleotides) can be lethal. On the other hand there are trisomies where an entire chromosome is duplicated. These may cause severe health problems but don't result in a different species.
As is discussed in many other questions on this site, the definition of a species is fuzzy and context specific.