This is only partly true. But SSD is just one piece of the puzzle, you also need fast decompression unit to go with it. PCs do not have dedicated hardware asset decompression units like the consoles. The temporary solution is to fall back to CUDA cores via RTX IO until there's a silicon implementation sometime in the future.
Reposting the same images of marketing buzzwords doesn't lend you any more credibility. A 10x faster storage medium doesn't give you 10x more performance, that is not how system performance works as there are many other factors at play here, shaders and graphics rendering being one major example.
Currently PC use CPU & RAM for decompressing/processing data before the data is accessible by the video card. Direct Storage will change it to Video Card & VRAM resources. In other words the data will go from your (NVMe) SSD straight to the video card for decompression & processing. That is it, that's the whole magic behind this API.
Ah, and here we go again. So Microsoft, NVIDIA must be foolish to develop and bring technologies like RTX IO, SFS, and DirectStorage to PC because... more memory... You could have literally saved them a ton of R&D cash and solved the problem right there, buddy.
Again, nobody is saying that Directstorage isn't beneficial, so stop it with the frikkin' straw-manning.
What people are taking issue with is the unsubstantiated hype and overexaggerated representation of this API by fanboy who do nothing else but lap up the sensationalist marketing talk. The hardware for this has existed for a long time on PC now. These developments are made because the massive console audience has now finally switched to SSD technology also.
That's not how it works. The pipeline between storage and system RAM is still massively slow and has bottlenecks. NVMe drives are nowhere close to being fully saturated with the current ancient storage APIs. Transfering huge amounts of data from storage to fill up all memory is still going to be painfully long if you don't solve the I/O bottleneck that DirectStorage is looking to solve.
That is exactly how it works. The CPU on your console has the exact same bandwidth as the CPU in your PC system, because they are both functionally the same. GPUs need higher memory bandwidth because their many cores are designed for parallel computing while CPUs need bigger memory size for performing serial tasks.
Nowadays, the best CPUs have about 60GB/s while the best GPUs have 750GB/s memory bandwidth. So offloading some of that workload from the CPU to the GPU is beneficial to system performance.
When consoles are advertising their bandwidth, they are talking about GPU bandwidth. Even an old 780ti has 336 GB/s bandwidth, which is exactly comparable to the series X.