I don't like this, but I can't say that it's much worse than, for example, advertising a CPU's single core turbo speed as generic "clock speed" which happens more than it should. Plus there's the issue that both the performance and system architecture of small Optane modules make it something that doesn't fit nicely into either of the data-storage buckets that the typical customer are familiar with.
This is worse because it's not Optane as memory, it's Optane as storage and they know it.
We're not talking about machines using Optane memory devices
here (i.e. this is not Optane DC Persistent Memory DIMMs
which exist and go in DIMM slots on a server motherboard), these are Optane storage devices
, 16 GB M.2 NVMe
These laptops e.g. Dell Inspiron 15 5570 typically seem to have a single SO-DIMM of DDR4 2400, which has a write latency of ~50 ns and write b/w ~19 GB/s ; their Optane M.2 16 GB device has a spec'd write latency of 30 μs and peak theoretical write bandwidth 145 MB/s
That review tested a 64 and a 32 GB version with spec of 640 and 290 MB/s write and they delivered at QD1 random write performance of 129.9 and 97.5 i.e. 1/5th and 1/3rd peak bandwidth respectively. Apply that to the 16 GB model and you get (being kind) maybe 45 MB/s
So, the ratio here versus actual memory
is 600 on latency and 130 to 390 on bandwidth.
i.e. 30,000 vs 50 ns (where bigger is worse) and
50-145 vs 19,000 MB/s (where smaller is worse).
In the case of those CPU turbo vs base clock speeds that you mention, it's nothing like as bad though, because they are the same thing and the effect is a factor of 2-3 not a factor of 600.
e.g. the i7-8500Y with 4.2 GHz 1T boost clock and 1.5 GHz base clock, and IRL the clocks will mostly be up nearer the 4 GHz than the 1.5 GHz level.