I doubt many pure mathematicians would be okay with totally subordinating all exploratory mathematical investigation to specific practical problems, and the field would look very different both in terms of how it's organized and what output it produces were it run that way.