With an Intel CPU like yours, the performance difference between 2666MHz CL15 and your current OC'ed 2800MHz CL19, is so small I doubt you'd be able to see the difference while gaming.
I used to have a set of 32GB Kingston 2666MHz CL 16-18-18-39 which I replaced with a set of 32GB Corsair 3333MHz CL 16-18-18-36 RAM, and I'm unable to tell the difference in performance, unless I use an FPS counter or run various benchmark tools.
I can measure the difference, but I can't actually see it with my own two eyes.
Although you can't compare our systems directly, I just ran a 3DMark Firestrike Benchmark for fun, with my RAM at 2666MHz and at 2800MHz (
same timings in both runs = 16-18-18-36)
Here's
a link to the online comparison between the two runs
Most of the differences fall within the margin of error, and especially in
Graphics Test 1 and
Graphics Test 2, the difference is so insignificant, that they can't hardly be noticed in real life, in my personal opinion - and at a closer look, it seems like the benchmark isn't even precise enough to accurately measure the difference.
As an added comparison, here
a link to the performance increase from
2666MHz CL 16-18-18-36 to
3500MHz CL 16-18-18-36 (+834MHz)
As you can tell, Graphics Test 1 and 2 still fall within a
0.5% difference with the
834MHz increase, and could be down to margin of error again.
To try an actual game, I also ran the built-in benchmark in Far Cry New Dawn, and at 1080p there was a 3 FPS increase (from 116 to 119)to the AVG FPS from
2666MHz to
2800MHz, while at 1440p there was a 2 FPS increase (from 106 to 108) in AVG FPS.
2 - 3 FPS isn't really noticable unless you are currently seeing very few FPS with your laptop.
From
2666MHz to
3500MHz in Far Cry New Dawn, there was an 8 FPS gain at 1080p in AVG FPS (from 116 to 124), which is a bit better.
SO - In short, If I was in your situation, I personally lower your current RAM OC a bit and see if you can tighten the timings instead (CL19 at 2800MHz is not impressive), or completely disable the OC, to be honest.
It could be interesting to see what your RAM OC would look like at 2666MHz and the
best Cas Latency you can achieve with your current RAM at that frequency.
(and compare that to the Kingston RAM you are considering to buy - and then reconsider the 100 USD investment).
But in my opinion, 100 dollars USD for the 2666MHz Kingston set, is too much money for such a small potential increase in performance - you may not even notice any difference, which I personally think is what is going to happen.
In a
worst case scenario, if your RAM run hotter when OC'ed, the heat build-up when using a laptop, could
in theory affect GPU and CPU clocks in a negative way. It is probably not a likely scenario, but there are sometimes negative impacts seen in realtion to overclocking.
As
Koekieezz has already mentioned, I honestly don't see a lot of potential for increasing performance much more than what you are already seeing with your current setup.
EDIT:
I chose the Firestrike Benchmark because it runs at 1080p and is less demanding to run, since you specifically mentioned Fortnite which isn't extremely demanding to run, and you have a laptop with a 1080p display.