FEA will only give you stress for a given input force/load. It is also dependent on how you model the joint. Clearly there needs to be a gap for assembly, so is there a glue or weld? That complicates FEA in how those stress risers are applied. Once you have a stress from FEA, you then need to know how the load is going to be applied over time, and then you need to know how the material will handle the load over time. While material properties of carbon fiber are pretty well understood, whatever bonding process is not. In my world steel plates are easy but as soon as you throw a weld into the mix it gets a lot more complicated.
Here's my attempt at a free body diagram for forces. If a 200 pound person is standing stationary, that 200 pounds would be divided between both offset pieces. Summing the forces in the Z direction, the 2 offset forces have to sum to 200. I just put them each at 100 for simplicity. Where it gets more interesting is if you some the torque around the tree contact point on the bottom offset. I get that there is contact area at the tree and not just a rotation point, but in dynamics and tree variation, you probably approach a rotation point. That would mean the joint would have to support the force the person is exerting on it during a climb.
Someone feel free to shoot holes in my attempt.
Sorry for the multiple pictures. First one was rotated so I fixed that, but apparently loaded multiple in the process that I now can't remove.