there are a number of blog posts that have different details about the how/why, etc. i just followed the links in the article to other parts of the series.
I expect that the use case is more prevalent than you think, where you are spending a decent chunk on cloud infra. I have been convinced for some time now that the costs are high compared to our on-prem. I really like the idea of a the “deft” type hardware management service, so that look after the DCs, hardware and connectivity, and we look after the software.