It seems like you didn't read the full report.
But the problem was, that I did, and I read what the dude said.
The company dealt in data. Forgot if it was backups or any of that but whatever.
How they choose drives follow the following:
They keep a close eye on the market constantly to buy new drives. When they see a few HDD's that are in their price range and general specifications, they buy them and run them through a certain benchmark test. If it passes or reaches their requirements, it gets added to the buy list and bought on mass.
NOTE: That the blogger, company writter noted that they only bought HDD's that fit the size, price, and specs that they needed. This could mean they bought a certain enterprise HDD or even not. Also note that they are buying a certain HDD of a size, and that is limited by price which means that you are getting a market of the HDD sector of that company limited to that size spec, and MSRP or street price aquired price. Also note that their benchmark test does not necessarily mean it was able to weed out the HDD's that won't last the longest. The strenous stress test their benchmark put some drives through just means that the drives that passed it would fit the conditions that they needed and set forth to want. Not the most reliable drive. You could have a drive X that in reality (by hidden Seagate numbers) would be more reliable than drive Y. Problem then being that drive X can't sustain the sustained and constant read writes, and temps that the benchmark gives and thus drive Y made it past. Then we must also consider factors again in which the company chose not just enterprise drives but just any and many drives that fit their size, spec, and cost sheet before benhmarking
The writter clearly mentioned that the graph and data was not conclusive to anything in the real world in terms of how Segate or WD etc perform. It was just data that their company had based on the drives that they chose which were based upon requirements that their benchmarks needed
This is the 4th time I wrote this. I'm not saying the data is useless. It's not, for companies like the one that did it or any VERY similar to it and that would use its benchmark, and aquisition pattern of HDD's, this data is invaluable. However, change a few of those factors (say that their aquisition of HDD's was different such that the price was much higher or lower, or size being different), and the data is no longer completely reliable for that sector.