one could certainly be evolved for the task in short order. the enzyme they used looks like it’s target is close enough structurally to use as a starting point.
you amplify the gene (PCR) under low fidelity conditions, and look for clones that decarb our favorite cannabinoids.
I’d wager an enzyme would be the most regio-specifc way to put the carboxyl back too. the best candidate from above might be a good starting point, under different reaction conditions, but screening through the first pool as well makes good sense.
If I could see a good way to make decarbing/REcarbing THC a life saving (and thus selectable) trait, it would be faster.
Edit: …and now we’re officially off topic. thread title is “Chemistry”!!