Wrong, wrong, wrong!
"first establish the baseline cost of following the “old” way, like staff’s pay, tooling, and the expected cost of fixing screw-ups"
...and then give the work to Developers who don't have years of experience in Ops, and only spend a very small percentage of their time focused on Ops resulting in way more screw ups to fix?
OpsDev would be a better idea because Ops deal with real production issues and know how it needs to work in the real world. In fact... I'm off to patent OpsDev. Or is it copyright.. Or should I release it under GPL?
How about the author arranges actual comment from organisations that have used DevOps successfully, noting IT estate size, number of staff involved etc... Because all I'm seeing in this rag is marketing bullshit, although given it's written by a "marketer" it's hardly surprising.