The games are not meant to be using comedy RAM, you're using 32GB to compensate for incompetent original or third party code.
Which means you do have 32GB for a novel reason.
Well to play a game isnt novel to me
Also with 16 gig of ram even with no game loaded I found myself having to restart chrome every now and again to free up ram, now I dont need to bother.
Granted the vast majority of people 16 gig of ram is plenty, I never said its a common requirement, just that for some people its not a novelty.
I agree, for the most part. Your chosen CPU, GPU and monitor should be well balanced so that they compliment each other well and perform together optimally. However, presenting a GPU limited benchmark chart as "evidence" for CPU performance is simply flawed. According to that chart the Ryzen 7 2700X is only 5% faster than a Ryzen 3 1200!
Low resolution benchmarks on high-end cards do not represent real world usage. But the intended purpose is to remove any GPU bottleneck from the equation so that your are only comparing the performance of CPUs. This is useful for projecting future performance; e.g. games which may be more CPU demanding and new GPU hardware which can take advantage of the the additional CPU performance. It's not just about getting insane frame rates in e-sport games.
It actually annoys me when I see things like only 4k benchmarks on say a ti card or developers forcing specific configuration to "ensure there is no gpu bottleneck". To me the system works as a whole, so e.g. when you bench/review a GPU you dont deliberately pick non cpu bottlenecked games.
It also annoys me if someone says e.g. if you buy a ti you shouldnt be 1080p gaming, these are all assumptions about what people want or do, and also assumptions that every game you play is well coded and optimised, the earlier example about disregarding my 32gig ram needs for FF15 is an example of that "oh the code is a mess, so its not relevant in terms of hardware requirements".
I agree with you on the balance, the balance is different for each person tho, and many reviews dont emphasise that either like say pairing a gtx 1060 with a 8700k. The excuse will be we dont want to have the cpu slowing down the gpu, in real life these things happen and tests should reflect real life, in real life people play horribly unoptimised games, tests should reflect that also, I also cannot remember the last time I seen a single or dual threaded game on a hardware review yet those games exist in large numbers.
Instead reviews are often done in a manner to paint the hardware been reviewed in the best light possible, hence 4k only reviews for RTX series ti cards, and using unbalanced hardware with the aim of ensuring the bottleneck is fully on the hardware been reviewed.