The new memory standard offers more bandwidth than the one it replaces, costs considerably more, and will make almost no difference to how your computer feels. I would skip it.
The reason is a distinction that gets flattened in every specification table. Bandwidth is how much data can move per second. Latency is how long you wait between asking for something and the first byte arriving. They are different properties and the new standard improves one while making the other slightly worse.
Which of those matters depends entirely on what the machine is doing. A workload streaming enormous contiguous blocks cares about bandwidth. Nearly everything on a desktop is doing the opposite: many small scattered requests, where the processor asks for one thing, waits, and asks for the next thing based on what came back. That is a latency problem, and adding bandwidth to a latency problem achieves nothing.
So you can pay a substantial premium for memory that benchmarks higher and behaves the same.
This is the same shape as the argument I made about megapixels on compact cameras, and it recurs because it is structural rather than accidental. The industry competes on whichever number can be printed on a box and compared without explanation, and that number is frequently not the one that determines the experience. Nobody is lying. The figure is accurate. It is simply measuring the wrong thing for the customer reading it.
What actually transforms a machine is quantity, and it is a deeply boring answer.
Going from one gigabyte to two, or two to four, changes the character of a computer, because the point at which the operating system starts moving things to disk is the point at which everything becomes glacial. Disk is thousands of times slower than memory. Every moment your machine spends swapping is a moment it is running at a tiny fraction of its capability, and adding memory removes those moments entirely.
That upgrade costs less than the premium on the faster standard and produces a difference anybody can feel from across a room.
Buy more of the cheaper memory. The faster kind will be worth having in two years when it costs the same as the slower kind does now, which is when every new standard becomes worth having.