what is the actual value gained from staging specifically? Once you have a list of those, a specific list, figure out why only staging could do that and not testing before or after. And "it's caught bugs before" is not good enough.
I believe modern day testing infrastructure looks very different. I have seen products like ReleaseHub that provides ondemand environments to dev to testing their changes out which eliminates the need for common testing env. That naturally means you need atleast one "pre-release" environment where all the changes are which would eventually becomes the next release. If you don't have this "pre-release" environment you will never be able to capture the side-effects of all the parallel changes that are happening to the codebase.
Thirdly, you have to see the context. When you have a microservice architecture, having a staging environment does not matter as fault tolerance, circuit breaking and other concepts makes sure that failed deployment of one services does not impact others. However, when you have a monolithic architecture you will never know what the side-effects of changes are unless you have a staging environment which would get promoted to production.
If you value customers, you should have a staging environment as a guardrail. The cost of not adhering or having a process like this is huge and possibly company-ending.
Why isn't it good enough?