You know, something else that came to mind regarding the descriptions above of +5-20% in this vs that etc, is definitely a thing that perhaps people should take more measure on how they actually use a PC.
Person A owns : Xeon E5 1660 v4, 64GB Quad Channel ECC Ram, SAS 6TB Array + 2TB Enterprise PCIe SSD etc.
Person B owns : Core i7 4790 non K, 16GB DDR3-2133, HD 7870 2GB, 240GB SSD.
Say person A is an independent media professional, who encodes dozens of hours of 1080p and 4k media per month, perhaps on contract, or perhaps for a YouTube channel etc. They want to build a single PC, and they make income based on volume of content produced. This person 'uses heavy multithreading' for sure. They also game from time to time when they have the urge, but maybe only a few hours a month at most, and they can't be bothered to buy a cutting edge GPU, and are happiest with 4k 60hz large format displays, and run a pair of them off a last gen Pascal or Vega card. Because of the volume of their primary heavy use case, +20% or more improvement in time efficiency is HUGE. They could conceivably take on more work, or be able to use their time more effectively, spend more time with family, hobbies, getting outside, whatever.
Now take person B. They encode a couple of times a year as the family member called on to help make videos of family reunions, quinceaneras, bar or bat mitzvahs, gatherings of the Hell's Angels, or whatever. But basically something they do one weekend two or three times a year, and maybe the odd Photoshop job a cousin messages them about over Facebook etc. They also "use heavy mulitthreading", but say you even were able to double their effective encoding performance, sure it might cut 20 hours a year down to 10 hours a year, but it's not going to really affect their day to day life in any real way, and if their PC is already good enough for their purposes, upgrading might be a bigger risk for them vs just riding it out a while longer or upgrading an area where they might see more benefit (say going from HDD to SSD, or getting a new monitor with better size and quality, etc). They game a few hours a week with friends on common MP titles like Apex Legends and COD Warzone.
Person A might see value even in investing in a new Threadripper 64C/128T rig, or multiple rigs, if the economic results magnify their revenue and profit vs expenses (to say nothing about potential equipment write-offs). Even 5% advantage when you're talking big numbers and actual improvement in quality of life is a real consideration.
Person B, ehh. They might get wrapped up in reviews and benchmarks, and buy something that really doesn't even make all that much sense. Particularly if they have a limited budget, and they could have made a much more QOL-focused purchase that made their experience nicer. Maybe their KB/M was trash, or they had an old 1080p 23" TN panel with trash color accuracy, or their 240GB Sata SSD pushing past 85% utilization so it's starting to really chug and lag, etc. Going from a 4790 to a 10600 or 3600 rig would basically do nothing for then unless they also spent at LEAST $400+ on a GPU. And at that point they'd also be needing new DDR4 Ram, probably a new SSD, etc. Could turn into a $800-$1k+ upgrade in a hurry. Or they could spend a fraction of that on a 1660 Super, 480GB SSD, and a 27" 144hz 1080p AOC Freesync IPS, and immediately their gaming time would massively improve, at half the cost, and be arguably a better experience than spending double the money, getting an entirely new platform, but still on a 23" trash monitor.
Just goes to show that the answers for everyone will be different, and just because they vary, doesn't mean they're wrong really, I mean I might have the opinion that people can more effectively hit their goals within limited budgets by doing more intelligent research and analysis, but at the end of the day : are they happy with what they chose? If so, I mean what can we really say, even if they chose something ludicrous haha.