The stress results from my model showed the highest stress was in the thin handle at the top, not around the hinge at the bottom, but that is because I assumed the user holds the part at the top. If that was the wrong assumption, and say, for example, the user grabs the arm around the thicker part near the rectangular hole in the middle, then the highest stress would probably be near the hinges. It's important to identify the worst use case and analyze that. If I made the right assumption, then the simple model I described above has adequate detail around the connection.
However, anytime there is a hinge, there is likely to be a worst-case load compared to simply operating the hinge as designed. What does the arm stop on, in either direction? How much force can be applied to the end of the arm when the arm hits the stop? Those load cases probably generate much higher stress levels in the part than the normal rotation of the arm against the hinge torque.
Bolt pretension requires some geometry prep work, but in this case, I don't think it will be helpful. The reason is that the keensert is level with the face of the part, so Bolt Pretension will only serve to squeeze the steel keensert against the aluminum hinge. Since we are not concerned with those parts failing, the bolt pretension doesn't provide a benefit.
To your question on cycles to failure, you need fatigue data for the glass filled Nylon to answer that question. This data is sometimes available at the material supplier's website, but most of the time I look, it is not there. Contact the material supplier and ask for it. It takes the form of a table of Stress Amplitude vs. Cycles to Failure and has many data points over a range of stress values. This will only tell you about a failure due to cracking.
Another analysis you can do is thermal cycling. Since the steel kenserts and the gf Nylon have different thermal coefficients of expansion, if this assembly experiences frequent large thermal cycles, a cyclic stress will be created at the interface between the keensert and the Nylon. That stress could be larger than the stress due to rotating the hinge against the torque load.