Since they are full of holes, it's not likely to have much torsional strength, and it would be a huge science project to figure out the capacity anyway. You're better off getting rid of the torsion.
If you aren't actually relying on the bending capacity at the corner, then go with the second detail and call it a day, no need to check development length.
I couldn't digest this entirely - but I ran my own check and also got the same answer except with 37.5 degrees being from the vertical line and not the horizontal one. I think that makes more sense - the wider the plate the closer it comes to just folding in the middle.
Some level of checking is baked into the allowable stresses for wood. Would that post be ok if it was sized for select structural? Maybe not. I would probably recommend the owner continue to monitor, but it's likely nothing.
I don't actually ever use it for slab design - just too tedious. Only ever for slab evaluation - once you know exactly the layout and the reinforcement, and only vary the loads, it's not as bad
It's really not a program, I just type in the equations (different depending on the yield line mechanism you are looking at), and just play with the inputs until I minimize the energy dissipated. This one is for a yield line mechanism described above on simple supports.
I recommend Practical Yield Line Design by Kennedy and Goodchild, you can probably find it online for free. I've never heard of BA's method of differentiating the equations for work or energy - I just plug the formulas into mathcad and iterate the inputs til I'm happy.
I ended up working it out - you do end up with more yield line in total as the diagonals extend, but the work ends up being UDL x area x delta/3, and in the area that is a rectangular instead of a triangular, the work is UDL x area x delta/2. So more yield lines with the triangle but less work...