In fairness, the ecosystem around Excel is more than 30 years old and the community's expectations reflect that. There's nothing stopping someone from writing a test harness for Excel files...it's all automatable. There's nothing preventing a person from storing Excel files in version control or running a diff on the underlying XML.

The reason it isn't done is because by the time end users care about those things enough to get around to implementing them, they're probably ready for a process that already addresses those problems with baked in functionality...e.g. an RDBMS and application layer. When there's a business case for something better than Excel there are products and services and consultants that serve that business case.