You accused me of ignoring two questions. So here was my answer to your question about why grade 8 can be to hard for most general applications.
Ironhorse74 said "Notch sensitivity increases as hardness increases."
Here is a paper on the subject
http://www-mdp.eng.cam.ac.uk/web/library/enginfo/textbooks_dvd_only/DAN/threads/fatigue/fatigue.html
In the late eighties I was involved in testing counterfeit bolts in the Army's inventory. All were Chinese and none passed testing. We hauled fastners to the scrap yard in dump trucks.
I think you need to reread your link.
The integral of the stress-strain plot and/or the yield strength are the relevant properties for a properly applied fastener that experiences minimal shear (proper use of a threaded fastener uses fastener tension to induce friction to hold two surfaces against shear). Regardless, fatigue strength is a factor of the percent loading of yield strength, independent of hardness.
Grade 8 fasteners have significantly higher toughness and yield strength values than grade 5 or anything below that. The common belief that they are more brittle has zero data that backs it.
Fasteners are tested and rated as an assembly. That's cool that you can cite some analysis of a section of a bolt under a microscope, but that is not representative of the assembly. I did the same tests for reports while I was getting my degree in metallurgical engineering. Very little of that lab data done at a university is representative of the real world, as I learned while working for a major fastener manufacturer here in the US.
Can you reference one steering or suspension application of a DOT approved vehicle that is held together with anything less than a gr8/class 10.9 fastener? Speaking of real world applications, that's pretty much the definition of cyclic loading, coupled with some significant safety factors.
