You missed it, it was said by Gene K that after 500mi it knocks down a grade. I was just making reference to that.
The oil itself does not shear, it's the little plastic strings that do. And yes, how much shearing is there? Some, maybe a little, maybe a lot. All shearing of the VIM's cause grade to decline. But each grade has a range. A little in same range (grade) is no concern, but moving down into next grade might be of concern? Maybe. A 0w60 falling down to a 0w40 might be normal, there's a ton of VIM's in a 0w60 vs say a 10w40 or a 20w60.
I believe everybody participating in this discussion is already familiar with the mechanism of shear, both temporary and permanent, so no need to spell that out. The debate was focused around the claim that shear was the primarily driver of viscosity loss and the assertion that this phenomenon rapidly lead to an oil going out of grade, which has been contested.
No, not all shearing causes the grade to decline. HTHS for example is measured under high shear conditions, but the condition induces primarily temporary shear where the VII polymers flatten out and their effect on increasing viscosity is reduced. Once they exit that environment they coil back up and function as they did before.
There's a similar discussion taking place in the additives forum on viscosity loss that got more technical than this one and I suggested the use of both a visc calc as well as referencing the flash point of the oil relative to virgin to discern if fuel dilution plays a larger role than indicated. This is primarily of value when discussing Blackstone UOA's as they do not directly measure fuel dilution via GC but instead infer it from flashpoint, I assume based on a "reference" virgin value that may not align with the lubricant actually being tested.
The example I used there was my own M1 0w-40 UOA from "back in the day" from my M5 where GC was used to discern fuel content:
You can see visc is at 11.24cSt and fuel is at 5%. Virgin for this version of M1 0w-40 was I believe 13.8cSt at the time.
Using Widman's Visc calc and 0.550cSt for gasoline:
www.widman.biz
If I plug in 95% 13.8cSt and 5% 0.550cSt I end up with a final visc of 11.55cSt.
Do I consider this accurate? No, but it shows the potential effect that fuel has on viscosity and at least somewhat aligns it with an actual GC measurement of fuel content.
My assumption is that as dilution occurs, because of course it does not happen all at once, some of the lighter fractions flash off while others remain to both reduce flashpoint and viscosity, but by less than if those lighter fractions had remained. This why 5% fuel as measured via GC does not have the same impact on FP and visc than if you were to just replace 5% of the oil with gasoline and measure those properties in a lab. This is why real world results of that level of dilution, which happen, do not have the catastrophic impact that were predicted by the figures discerned from laboratory testing.