Interesting! Might play an important role in my decision to purchase an HDD since i seriously need one.
Our experience is fairly opposite from BackBlaze's WD vs Seagate, but it's dead-on with HGST-Hitachi's. I'm uncertain how big their purchase-lots are, and that'd be interesting to note, so I've asked. But our distribs maintain much larger records and they don't do "purchase date/lot" segregations which I'd think a assembler-manufacturer would really want to know. Or maybe that info's just not public. There is an interesting point - "Why not buy 2 cheapos that may not last instead of 1 Enterprise?" What they don't mention is the consumption of electricity, which over a few years is a large Dollar Amount, too. But that's a pro-SSD argument. One thing that's curious - WD 3Tbs are numerically listed at 8.8% fails, but the graph's bar isn't close to the 8% level. Odd - there seems to be some other mis-drawings of bars as well, or else a miscalibrated X axis. EDIT: Yes, I think "miscalibrated X Axis" is the likely culprit IF you follow Numbers to Bar-Chart highpoints. It appears the relative scale among them is correct.
It suprises me that the its more cost effective (including labour effort) of using consumer drives. We get 1st class service for replacement and new drives from HP/Dell, just one email and we get a new/replacement drive next day delivery. Do blazeback even bother using the consumer RMA process ? We aim to keep hardware for 5 years.