Thanks for experimenting with me @HyperJet2018 and @RPthreenine. @Nickba78 it looks like you shouldn’t keep trying as we’ve established that it does not make a compelling difference.
As I mentioned earlier, I wasn’t convinced this FOV change would make a difference, and I’m not sure I understand your point on “render Resolution did not change, i thought that it gets decreased”. The resolution is another parameter (the Custom Render Scale, which you are already playing with). When you changed the FOV to be smaller, less pixels ended up being rasterized because the projection did not cover the entire screen. This can save up on frustrum culling (the ability for the engine and geometry pipeline to skip drawing geometry outside of your FOV) but for such small reduction in FOV I don’t think it would be noticeable. Less pixels can also save up on the pixel shader stage, which for some applications can make a noticeable difference if they have very intense pixel shader processing, but it does not look to be the case for MSFS. When RPthreenine tested with 50% and saw a bump in 6-7 FPS, this was likely the combination of both savings described above, because the FOV was significantly smaller.
Anyway, in spite of the results not being what you had hoped for, I am glad I could assist with this experiment. I hope it also gives you a good example of why not all feature requests can be addressed 
For C/C++ developers out there, I reiterate that you can use the mostly-generic API layer code I shared to conduct this kind of experiments. Maybe with luck you will find that “game changer” tweak
PS: To the mods, I apologize for hijacking this thread, given how many messages we have now, I wish we had just started another thread! Sorry.