If a single point of failure took down your entire network, then your solution wasn't engineered properly. Just because on provider messed up big time doesn't mean every other provider would. The hardwdare design/architecture can also be provided to you for free as part of the sales process, no extra service required.

As for hardware failures in general, you can afford to have some complete extra servers on hand doing nothing most of the time with the amount of money you'd save on renting bare metal instead of paying for cloud. Or if engineered properly, you'd be using all the machines at once but be able to afford to lose half of them, and failover automatically.

Yes, a cloud (bunch of VMs on a cluster) is going to be more reliable than a single machine, as you get hardware redundancy. But a cluster of machines with hardware redundancy is going to be much more reliable than a bunch of VMs on a cluster, as there's less complexity involved without the cloud stack, and thus fewer points of failure.

In our case, I mostly agree - we however were led to believe we had network redundancy and when they added new servers and it effectively blacked out access to all servers we realized this was not the case. My realization at the time was simple. It's better to be on a ship with a lot of engineers dedicating their time to a bigger system then hoping an engineer finds the time to check on your one off raft managed solution.

Public clouds are an efficiency gain for both the datacenter and the end user. The bottom line to me is more engineers are focused on operating and improving the public cloud than the state of the art managed dedicated racks from 2014.