Very early in NASA's Mars landing efforts, one organization was operating in metric and another in English. Without knowing that, calculations were made and the descent was started.
The landing completed, but the moon rover sort of disappeared into the Martian soil. I'm sure you guess the rest of that story.
Data interoperability is about making sure those types of "events" never happen again. In data it's not just the data types and value domains that matter. It's also the meaning of the data fields, when values change, by whom, and with what level of accuracy. The factors are almost too numerous to enunumerate and at the same time, too critical to ignore.