I think one of the major points in the viscosity discussion is the actual temperature of the oil in operation. If the oil temp stays steady, then as the guys towing with 0W20 have seen, the oil will perform just fine.
But if the oil temp is slightly higher, then that oil is slightly thinner. So, if I knew the oil would stay at say, 90C, then I would be fine with the -20. But if the oil under hard use gets up to, say, 105C, then a -30 will offer close to the same film thickness at that temp that the -20 had at the lower temp.
What was really interesting about having an oil temp gauge on my Volvo (I know, another Volvo story from me) when I had it in Colorado was the relationship between oil temp, coolant temp and outside air temp. It wasn’t what I expected, and, I would surmise, what other people think.
The car had a turbocharger (initially oil cooled, later, upgraded to water cooled) and a large, external, thermostat-controlled oil cooler adjacent to the radiator.
In the winter, with weather below freezing, even below zero, the oil took a long time to warm, but it always ran at about 85C, the thermostat opening temp.
In the summer, even when the temperature was close to 100F outside, the oil ran at 90C. Not much of an operating temperature difference, really, despite the huge variation in outside air temp. Different than the assumptions most people make, and clearly the result of a good oil cooling set up from the manufacturer.
But, and here’s my point about being able to monitor temp with a gauge, no matter the outside air temp, when I used the engine hard (climbing up I-70 to go skiing, for example) the oil temp would rise up to about 105C. Part of this was the nature of that engine, with the turbo putting a lot of heat into the oil. No matter how they’re cooled, turbos put significant heat into the oil.
I saw nearly 115C when pulling a big trailer with the car, in the humid summer heat of Virginia. It was on Route 13, on the Eastern Shore. Lots of stop lights, reducing air flow over the oil cooler, and lots of heavy throttle acceleration to get the car/trailer up to speed.
There was a significant change in oil pressure across the RPM range with that range of oil temps. I really do wish for better gauges to include oil temp and pressure.
In the absence of gauges, in the absence of knowing how hard use (again, trailering) affects the oil temp of the engine in the Tundra, going up one grade provides a bit of “headroom” should the oil temp climb above normal.
The language in the owners manual says exactly that - it says, “The 20 in 0W-20 indicates the viscosity characteristic of the oil when the oil is at high temperature. An oil with a higher viscosity (one with a higher value) may be better suited if the vehicle is operated at high speeds, or under extreme load conditions.”
So, yeah, on the Tundra, I run 5W30.
It‘s cheaper, easier to find, and covers that potential higher oil temp situation provided for in the owners manual.
Now, I run exactly the specified viscosity in the half dozen other cars, all with turbochargers, because their OM viscosity specification doesn’t say anything about going up a grade for hard use. They all have oil coolers, too. The Volvos are coolant/oil, while the Mercedes are external air/oil, but they’re all fitted with oil coolers.