I feel we need to split this discussion in two as it greatly affects the performance implications being discussed here.
1) Load times. The time tolerance here is much higher in that you aren't likely going to notice load time differences in seconds and therefore performance differences between SSDs that could impact this in terms of seconds isn't likely to be particularly relevant in practice.
vs.
2) Streaming in data during game play especially per frame. Keep in a mind a single frame at 60 fps is just 16.7ms. This means potential performance differences between SSDs that might result in millisecond differences will matter.
Reasonable point, but I think people can and will adapt to a lower capacity than 2TB if they need to in order to play the latest, technically impressive games. It's not like you cant still have other drives, including any ones you've already got installed. Obviously my $60 example is a bit extreme as I dont expect many people will be happy with a 500GB drive for games, I was just demonstrating that this isn't like being asked to buy some expensive new cutting edge CPU or GPU or something. It's a relatively affordable, easy upgrade to make, using a slot that otherwise wouldn't be used for anything else(meaning you make no compromises in using one).
Even if there is a bit of a moan from some, the alternative of just 'missing out' when your PC is otherwise up to snuff will be too much for PC gamers to bear.
But there is a bit of chicken and egg problem here in terms of the incentive to upgrade and therefore driving adoption vs. incentive of developer support dependent on adoption.
Right now is there even a single game that will show very tangible benefits? If you say buy a $500 or even more expensive graphics card for instance we know tangible what benefits you will receive from that and how wide those benefits will be in terms of support. Compared to even a $60 outlay for a NVMe drive, over what they currently have, which may or may not be the right choice once the games come around actually "requiring" one.
For example take this posts -
On another note, now that I finally have a motherboard with an NVME slot in it I'm thinking doing I order an NVME drive now or wait?
Do I get a cheap 2.5GB/s one and upgrade later or just get one to match the Series consoles?
Hmm..
I would have a different opinion against some of the responses above in terms of upgrading now. Why would you upgrade now unless you expect NVMe drive prices to shoot up (NAND prices do cycle, so it possible we see prices rise sometime in late 2023 or later depending on the cycle and macro conditions, but this is another complex topic)?
We have no data on what the situation will be at all in practice once Direct Storage, or whatever solution developers go for, actually ends up being widely implemented in the wild. So I don't see how a person can be confident how much of a benefit they will get from upgrading. Also what upgrade makes the best sense. For all we know once implementations hit we do find certain drive characteristics do lead to better or worse real results. Or even the cheapest off brand QLC SATA ends up being the same. Who knows.
Forsaken is the first Direct Storage game still scheduled but delayed to Jan 2023 I believe? Even with that we won't know how representative it is for other releases with a sample of one.
