Where I think the nvram will help most is battery performance when you do longer term tasks.
For instance just chat with someone, write emails or surf a bit. What happens normally then is that the browser cache, email program or many of the commercial Instant messengers keep writing to the HD so it can never fall asleep.
Same goes for word autosaving as well as many other applications.
I especially see this problems with laptops that has been used for a while and has got lots of applications installed. There is always some apps that install small auto update check applications or some IM that keeps downloading advertisments or something similar that always causes some I/O every few seconds. This happens even if you do not actually do anything on the laptop and can be a real battery eater.
Read I/Os does not matter. They get served from the OS buffer cache. The writes does however as these has to be flushed to disk either in a few seconds after writing or immediately when the application decides to flush them or when it closes the file.
I have worked with IS related tasks for 16 years, and I have hardly ever seen a laptop used for more than 6 months by a fairly active user that install apps outside of the basic stuff that actually have HD suspension that still works as expected. The HD tends to spin forever or at least much longer than it should.
I would think the NV ram write cache would help significantly on this problem, but 32MB seems very small.
I suspect that the benchmarks done here does not really reflect this as they are probably used with a 100% clean OS without any "bad" (for HD suspend) stuff running, which would give the non-NVram equipped HD an advantage.
Also, the pinned/superfetched data will not help much in typical benchmarks. Where this will help is on system responsiveness when doing other I/O tasks that would normally cause applications to be flushed from OS cache.
I am sure it will help there, but only if the memory on your laptop is fairly small.
I use a fast 4GB USB memory disk for readyboost on my desktop which has 4GB memory. Anything less than 1GB does not give any effect and the effect with more USB memory is mainly when I do I/O intensive stuff. I frequently process data that are many times larger than the 4 GB I have memory for, and readyboost noticably helps on system response when I do that.
I also notice a lot less seek noise from my Raid system (which is good since I have spent a fair bit of time and money to get a silent computer). I actually immediately notice it based on the seek noise if I pull the USB flash, but the raid is so fast that when the system is not loaded, there is very little difference in performance. There is a tiny bit more "snap" to some tasks when the readyboost is enabled, but it would make no real difference to my own efficiency.
Such benefits are however quite difficult to quantify in a benchmark scenario. I see it clearly with my own eyes that it helps in some scenarios that makes it worth the investment for me, but few benchmarks, if any, today will reveal this.
I do not expect people with >1GB ram to notice much difference on daily tasks.
I would have wanted to see more memory used for the write cache (64-100MB) so that more could be written without spinning up the disk and the OEM/superfetch area a bit smaller as they will not help much if you have more than 1GB memory, and I expect most laptops that will include HDs like this will be pretty well equipped in terms of memory already.
I suspect that the benchmarks done in this review does not reveal the real benefit that I think will be from the NV ram write cache. I do not think this will show until you have a 6 months old laptop with lot of "trash" installed that keep the disk from suspending itself.
Then I would not be surprised to see 20-30 minute extra battery life on light tasks (browsing/email/etc) if the NV ram write cache works well.