I'm curious how the data stacks up by SSD brand. Really in SSD quality in general. Such as the memory controller used or type of memory used.
Going by brand might not provide a fair comparison, as they may only be buying certain models within each brand's lineup, which might not be indicative of a given lineup as a whole. And even if they listed the drives out by model number, it's been becoming increasingly common for manufacturers to swap out the controller, memory chips or other components without giving a drive a new name.
I wonder if the M.2 card drives are more (or less) reliable than a standard SSD from equivalent tech from the same manufacturer? I'd hazard a guess that not being inside a case, they have better heat dissipation and could be more reliable.
On the other hand, M.2 NVMe drives will generally be transferring data faster, which can result in higher temperatures. And those not fully-covered in heat-sinks may be subject to dust-buildup on their components over time. As far as temperatures go, a lot of that may depend on where the drive is installed in a case too. A SATA SSD wedged behind the motherboard with little to no airflow will experience higher temperatures than the same drive in a standard hard drive rack with air circulating over it, and M.2 drives can also potentially experience differing temperatures depending on whether they are installed in a slot behind a graphics card or out in the open. Of course, one drive operating at a higher temperature is not necessarily indicative of something that would affect durability, as long as the components are not operating near their temperature limits.
We also lack details about what sort of operating conditions these drives are under, and how comparable those conditions might be to a typical desktop system. Both in terms of things like temperatures, as well as what sort of write loads they are undergoing on a daily basis.