1. No, it does NOT perform on par with a 1070. It is CLEARLY SLOWER THAN A 1070:
Proof:
https://www.vortez.net/articles_pages/zotac_gtx_1660_ti_review,8.html
Firestrike is a benchmark, not an actual game, and it is only one example of performance. Due to architectural changes, the 1660 Ti will perform faster than a 1070 in some games, and slower in others, but overall their performance will be very similar. Even that very review you linked to shows that to be the case when you look at the actual game results...
https://www.vortez.net/articles_pages/zotac_gtx_1660_ti_review,17.html
In fact, going by their results, the 1660 Ti was actually the faster card in most of the games they tested, though six games is arguably not an adequate sample size to precisely judge overall performance by. How about the average performance across 21 games as tested by TechPowerUp then...
https://www.techpowerup.com/reviews/EVGA/GTX_1660_Ti_XC_Black/28.html
Or the 33 games tested by Techspot...
We're following up to our GeForce GTX 1660 Ti review with an even more ample 33 game benchmark test. The day-one review looked at more recent games...
www.techspot.com
Pretty much every single review I've seen has shown the 1660 Ti to perform at roughly the same level as a 1070 across today's games. And if anything, the 1660 Ti's updated architecture seems to handle newer games better, which could potentially help it pull ahead a bit in the future, though the additional VRAM of the 1070 may help counter that in some games.
2. A 3570k is NOT on par with a Ryzen 3.
In fact, a 3570k overclocked to 4.4 GHz is SIGNIFICANTLY FASTER IN SINGLE THREADED PERFORMANCE THAN A RYZEN 5 2700X:
It is in FACT SO MUCH FASTER, than comparing them is
NOT EVEN FUNNY.
A 3570k@4.5 GHz does around 2600 CPU marks at Passmark Single Threaded whereas a Ryzen 5 2700X @4.3 GHz does 2194 CPU marks.
https://www.cpubenchmark.net/singleThread.html
A Ryzen 3 is SO MUCH SLOWER than an OC'd 3570k that comparing them is a
WASTE OF TIME.
Again, you're comparing a synthetic benchmark, and modern games are not single-threaded. Ultimately, there's not going to be a particularly large difference in performance between these processors. When it comes to more heavily-threaded games, things like Shadow of the Tomb Raider, Hitman 2 or Battlefield V, the higher clocks of an overclocked 3570K might help a bit, but the core deficiency compared to a 6+ thread processor is going to cause some performance drops, and this is likely to become more common down the line as games make heavier use of additional threads. And since you like synthetic benchmarks, let's see how a 2200G compares to a 3570k in systems tested on UserBench...
https://cpu.userbenchmark.com/Compare/Intel-Core-i5-3570K-vs-AMD-Ryzen-3-2200G/1316vsm441832
It looks like virtually identical performance at stock clocks, and the i5 can be around 7% faster with both processors overclocked. How can we even compare such massive performance differences? : P My point stands that these processors are in a similar performance range. The 3570K might be a little faster with an overclock, but not by enough for there to be a visibly noticeable difference in gaming performance.
3. The games that benefit from more than 4 Cores are SO FREAKIN FEW that I haven't even used the Hyperthreading function on my 4770k for the PAST SIX YEARS (I have it switched OFF) and there is no chance that I am going to use it either for the next couple of years either.
I agree that most games don't yet see much of a performance difference from moving to more than four cores (assuming other applications are not running in the background while gaming), but if you have had hyperthreading disabled for the past 6 years, then how would you know if it has an impact on performance in recent games? And why pay more for an i7 if you are effectively turning it into an i5 by disabling its biggest feature? I suspect that enabling hyperthreading would be better for minimum frame rates in many recent AAA titles than keeping it turned off to maybe get a slightly higher overclock.