>>why anyone does business with them, but they are likely to be around for a long time
The answer is partially in the question.
For large enterprises where large part of IT is a strict cost centre - the "boring" areas of HR/Pay, Financials, CRM, ERM, SCM, etc - predictability, support, longevity, reliability are far more important than sexiness, new technologies, or even immediate cost of license compared to cost of change and uncertainty.
And most Oracle "customers" probably arent interacting with Oracle directly. They are buying a financial system or erp from a vendor who is Oracle under the hood. They might get to the product selection stage and be looking at 3 industry specific products built on Oracle vs NetSuite. Either way Oracle wins.
I mean, yes and no. "No", for the purposes of many traditional non-IT large enterprises.
CIO of large company sees value in having "one neck to choke" for their HR Application, Proxy Server, Web Server, App Server, Database Server, Enterprise Service Bus, Search Engine, Monitoring Software, etc. A simple licensing and support model with streamlined contract model as opposed to 27 bespoke contracts with "never heard of them" consulting/support companies for each level of the open-source stack which "nobody owns" (https://www.postgresql.org/support/professional_support/nort...).
At ground level, we may not see value in the same way; we may complain and moan about it; about vendor lock-in, license cost, slow pace, etc etc etc. I certainly did, for decades :).
The answer is partially in the question.
For large enterprises where large part of IT is a strict cost centre - the "boring" areas of HR/Pay, Financials, CRM, ERM, SCM, etc - predictability, support, longevity, reliability are far more important than sexiness, new technologies, or even immediate cost of license compared to cost of change and uncertainty.