> Yes, that's what the article is about.

Not really. Hand-waving over hypothetical unit costs means nothing. In fact, I wonder if that's why the article lists no specifics on the cost structure of the spot instance based approach, and why the cloud 101 approach is spun to be such a surprise.

> And yes, the comment is confidently wrong as said in that page.

You're literally trying to gaslight when all it takes to verify your claim is a single click. Wild.