Agreed. I think the author took the wrong lesson, while you are sensing the real deal. There was a lot of rust advocacy time that ultimately had little clear business value -- a couple of bugs that may have been prevented, and rust may have cost different problems instead. If you're getting paid 6-7 figures and supposed to supporting 2-10X that in revenue, that's a real responsibility.
When an infra investment makes sense (lang, DB, ...), especially for 40+ engineers and who knows how big a workload, the multiplier should make the benefit clear. GPU computing speedups, k8s for easier elasticity, something, and ideally obvious like some sort of 10X. We certainly have quite a queue internally like that ("platform X unlocks property Y that has lift Z"), as do many. Conversely, with time wasted, that's a person distracting themselves and everyone else. Getting aligned with the business isn't about pushing a language that might help but identifying what is the most thing you can do to solve the company's/customer's problems, and going for that one.
When an infra investment makes sense (lang, DB, ...), especially for 40+ engineers and who knows how big a workload, the multiplier should make the benefit clear. GPU computing speedups, k8s for easier elasticity, something, and ideally obvious like some sort of 10X. We certainly have quite a queue internally like that ("platform X unlocks property Y that has lift Z"), as do many. Conversely, with time wasted, that's a person distracting themselves and everyone else. Getting aligned with the business isn't about pushing a language that might help but identifying what is the most thing you can do to solve the company's/customer's problems, and going for that one.