The author doesn't grasp what putting all your eggs into one basket means:

> Before anyone says “you put all your eggs in one basket,” let me be clear: I didn’t. I put them in one provider, with what should have been bulletproof redundancy:

That's one basket. A single point of failure. "But it should have been impossible to fail!" Backups are to handle the "impossible" failure (in reality nothing is 100% reliable).

This one time, traveling through Asia, a simple merchant transaction triggered a fraud alarm on my card. The default for my bank at the time was to cancel my card automatically. This was before the days where cards could become unblocked. I had to travel to another city to pick up a new card in 10 working ways. This was a Mastercard credit card. I thought I was smart traveling with both a mastercard and a Visa card. Well, the Visa card was automatically cancelled too. Due to the same event. No cards for me to use to get to that city and I had to resort to a dodgy western union transfer to move forward. Also, try booking a flight with cash, it’s not fun.

My point is that the basket that eggs are put in is not always clear in hindsight. I wasn’t even aware that Mastercard and visa shared fraud alerts and that they were automatically linked.

The author’s article is not about backups, it’s about accountability.

Fully agree. It's the same reason why you wouldn't put a repo on Github and then mirror that repo to Github. At a minimum any good mirror would be on a different provider, and ideally (if we get real picky) on a completely different cloud service.