My understanding is that this is what gain of function research is all about. Accelerated evolution within a lab towards infecting a specific species.

It doesn't have to be crispr introducing reach mutation.

RaTG13 is decades of evolution away from SARS-CoV-2, all over its genome.

You don't get there by splicing an ACE2 spike onto an RaTG13 backbone and passing it through a dozen mice. That gives you something that still looks similar to RaTG13 and infects mice.

The ACE2 spike also looks most similar to a previously unknown ACE2-binding spike protein found in malaysian pangolins.

So WIV would have had to have discovered that pangolin spike-protein, kept it secret, spliced it into an RaTG13 backbone, then not used mice but passed it through a species like that had a human-like ACE2 for a decade and millions of animals.

An alternative hypothesis is that Charles Darwin did that experiment.

That's fair. I wouldn't say that it's "we can't figure it out" but rather that "doing so takes a lot of mutations and thus resources that may be difficult to do under the table"

Sitting in a cave somewhere implies not being studied, right? Which is more the "natural mutation" thing.

Sitting in some government lab is a different story and depends on belief in scientific institutions to do science and publish peer reviewed papers and all that jazz.

Another question is how you can make a virus evolve in the direction you want. 20 years of bat evolution would not necessarily bring a furin cleavage site, so how do you make sure it does?