When it comes to SSDs, reporting numbers which are completely wrong or right only in unrealistic circumstances is quite common.
Do you know any website the staff of which welcomes any corrections and is willing to work to confirm and document them?
Sure, most of them around kind of want to correct it, but in the end don't because it is a very small aspect of their testing regime for which they have to essentially completely overhaul their test suite. It then depends on their quest for science if they agree to not publish these results at all, or keep the erroneous results on the site. Many really don't care about the difference, even if it is an order off.
By the way, Cryoburner, DIPM is only off by default on Windows XP and earlier (including Server 2003). Linux support is touch and go, and most test suites are either based on Linux, or slipstreamed/automated Windows Trace benchmarks. Also, DIPM isn't the only thing that isn't properly supported in these testing environments; generally, a lot of relevant driver features simply don't work causing very different power consumption figures across the board. I've seen same-generation CPUs paired with identical GPUs that had a 'system power consumption' of over 150W in one review, and just under 50W in another. The impact of drivers and the specific power management features enabled by those drivers is just very much unknown to the average reviewer (journalist). Hell, even the impact of the power supply rating is unknown to most.
The reason I'm so adamant on SSD power consumption errors is that not a minority or majority, but simply ALL reviews are wrong. The only way to find the lowest power consumption SSD is to go by manufacturer specs, and those are often incomplete.