Of course you could also invest heavily into cloud automation and likely run their payloads autoscaling and save a fortune too.
My point is, when people compare cloud to on prem they use a hypothetical on-prem installation vs a realistic actually working cloud deployment.
We only see these blog posts for things that are just 1-2 servers.
Very few companies are fully on-prem and saving a lot of money, they typically have very specific use cases like high bandwidth or IO usage.
Also, 6000 CPU "cores" on the cloud is more like 3000 CPU cores. Which you can get in just 20-50 servers. This is in the range of something that could be taken care of as a part time job.