Find an endogenous retrovirus identically positioned in the human and gorilla genome, and not present in the chimp genome in order to do tremendous damage to the theory of evolution. Find a few more similar anomalies in other genomes to bang the nails in evolution's coffin. Some creationists predicted that molecular evidence like that would falsify evolution as we mapped genomes. They were wrong. Instead, the retroviral evidence has stunningly vindicated evolution, matching as it does the predictions of the phylogenetic tree.
Alternatively find a rabbit fossil in pre-cambrian strata. Untold millions of fossil finds made since Darwin's day match the predictions of the theory of evolution.
Find an identical species of flightless bird native to two remote oceanic islands to falsify the theory of evolution. (Hint: we've already looked, and we couldn't find any, flightless bird species on oceanic islands are always unique to that island, for an evolutionary reason that you'll work out if you think about it)
There are loads and loads more ways that evolution could potentially be falsified, but where the real evidence when studied ends up vindicating evolution. Google is your friend. :)
No, he won't. That is the amazing thing about creationism/ID. It's the science of not thinking, not seeing, "I'll never get this and you can't make me."
I understand that considerable genetic differentiation between humans and other primates have been found. It has been found to such an extent that the smooth evolution theories have been problematized.
Nonetheless, it seems that the evolutionists simply respond--
its more complicated than that and dismiss questions as assertions from unreasonable fundamentalists.
I have to say my faith in evolution is declining with the noxious treatment of ID proponents.
That's a valid method of potential falsification in general, but in the particular example you chose, it would be a bit problematic. The reason is that the points of divergence of the gorilla/chimp/human lineages are close enough together in time that the divergence was more like a three-way split than a sequence of clean successive forks. As a result, it wouldn't be out of the question to find genetic "markers" that are common to different pairs of these lineages, in "contradictory" ways.