My point is that you can't generalize the behaviour of a Samsung SSD over to Toshiba or Sandisk. The latter have quite different implementation on the inside. It's not just Sandforce, there are differences between Sandforce implementations too, which is practically always neglected. And it's not about the NANDs, it's about the chipset. Low-level reformatting does not increase the lifespan of every NAND, of course, but it increases the health of the drive as the access table is being re-shuffled again, a known problem with some brands because the implemtation was known to produce inconsitent access states over time. Theory does not always work as expected. That's exactly what the linked article is saying. We'll meet again in a few years.