When I bought my '77 Camaro 305 (new), I knew I wanted to use Castrol GTX in it. It got 20W-50 GTX. I saw no real power/response loss with the heavier oil, nor did I look for it. On the coldest winter days (about 15 degrees F), there was one valve lifter that would "tick" for the first revolution or so, then quiet down. At about 300K, I started to use syn Castrol, still at 20W-50, then Rotella T 5W-40 for the higher zddp.
Somewhere in the '90s, GM put out a TSB to NOT use 10W-40 or 20W-50 in their engines. 10w30 was the then-current recommendation. I asked our Pennzoil rep about that, considering I'd already passed 150K miles on the Castrol GTX 20W-50 with NO oil consumption issues (as the TSB mentioned). His reply was that in some of GM's testing, they'd found "some" 20W-50 oil(s) where the viscosity improvers built-up behind the oil rings, leading to higher oil consumption . . . but it wasn't Castrol. So, as they found ONE oil that did that, plus that they didn't have their "Tier 1" oil list back then, they had to recommend against ALL 20W-50 oil use in their engines.
The OTHER side of things is that the particular oil viscosity they use to run their emissions and fuel economy testing is what they MUST advocate the owners use to ensure compliance with various laws (emissions, most probably) pertaining to their vehicles' performance and continued compliance with said laws.
Over the life of the 305 in the Camaro, it used 1/2 quart between 4K change intervals. This later became 1 qt/4000 miles somewhere around 200K or so, and linearly-increased from there. When it was replaced with a 350 that had been built for it, being stored MUCH longer than ever anticipated, the initial fill was with Valvoline 30 (the installer's judgment call). After 6K miles, I changed it to GTX20W-50. I noticed an immediate slight loss in throttle response and a slight decrease in fuel economy. I ran it a thousand miles or so just to make sure, ten put GTX 10W-40 in it and the throttle response returned, as did the fuel economy. No more 20W-50 in that engine! With the later concerns with SM oils, I went to Rotella T 15w40 and later syn 5W-40 Rotella T. Again, concerns about flat tappet cams and lowered zddp levels in SM oils.
When the whole zddp issue surfaced, one major aftermarket cam manufacturer recommended either Rotella T OR synthetic motor oil as "the cure". That was way back then!
Engine machining is much more "developed" now than in prior times. Bearing tolerances are still related to journal size than oil viscosity, I suspect, but almost all piston rings are now "low tension", so scraper design is more critical. On one of our LS-series GM engines at work, it saw mostly highway miles (250+ highway miles daily), I normally ran it to 10% on the Oil Life Monitor, which was usually about 10K miles. ZERO indicated oil use, at 10K and also the several times I ran it to 15K miles! On the recommended 5w30 Mobil conventional oil we get as a part of the GM Oil Program, at the dealership.
GM's motor oil purchase program has always been aligned with Mobil and Mobil 1 oils. Ford probably has similar programs, I suspect, but not sure about Chrysler. End result is that if the OEM is recommending a particular brand of motor oil in their higher-perf models, that's usually the brand the OEM's oil program is aligned with, by observation. Used to be that "cheapest" is where the purchases went, but with dealerships getting "credits" to purchase the OEM-recommended oils now, plus the many financial incentives dealerships get from the factory to participate in their many programs, no reason not to use the OEM-aligned lubricants "in bulk", purchased from the larger vendors who sell them to dealerships.
When the first "0W-__" oils came out, it was noted that such oils lubricated the valve train something like 30% quicker, on initial start-up, than even the 5W-___ oils of the time. This means better cold-start emissions and less wear as the lube system gets things done quicker with the lighter oils. I questioned that with our Pennzoil rep, knowing that the 0W oils would or should be the same viscosity at 40degrees F, I suspected. He noted that the 0W oils poured quicker at room temperature than the 5W oils did at the same temperature. Not much, but enough to notice. Be that as it may.
I know that some like to not comply with factory recommendations, for various reasons, but unless there might be a particular reason, a "real" reason than a "perceived" reason, then factory recommendations should be the default mode (related to climate, typically, as in the past). Plus the many international oil specs for emissions controls longevity. Even whether or not to use syn oil . . . which is now OEM-spec for some brands and engines! More variations and performance requirements which must be adhered to for warranty compliance and best engine longevity.
In the owner's manual for my '70 Dodge Monaco, 20W-50 is listed as "32 degrees F and above", just like 30W motor oil. 20W-50 is also listed for HD/performance uses, too. In that engine (383 HP), the engine would not like getting past 4000 rpm, surging, but this stopped with Castrol GTX 20W-50. It was not a fuel problem, but an oil pump "loading" problem, as the distributor is run by the oil pump, what they called "spark scatter" back then. In the middle 1970, Castrol 20W-50 was generally available only at motorcycle shops, being termed by many as "A motor cycle oil". Back then, too anything past 10W-40 was termed a "super oil" with the wider viscosity range and better additive packages, which were all very high-tech for their time.
Many might not consider the power consumed to run the oil pump, but it's there. Lighter oils take less power than thicker oils, which related to how much power is at the engine's flywheel (another "truth-in-advertising" issue, these days!!). Less internal power consumption can also mean less total fuel consumption, too, even if its barely significant other than in throttle response (when the "old" manual throttle actuation is used!). In more recent times, it's the base oil and additive package that tend to better define oil performance than just the viscosity rating, unlike "back then" when it was mainly viscosity rating.