When your dev teams are strewn all over the world and the lowest common manager with the people managing the OS is the CTO 6 layers up, you will find making things "just work" get exponentially more difficult and all sorts of workarounds and safeguards have to be put in to ensure small changes introduced by other teams do not screw up your portion of the app and vice versa, eg not letting the OS team control the JVM for fear security and performance decisions by the OS team will cause unexpected deployment problems... which will lead to performance and security problems that need to be worked on without OS team involvement...

Of course this is a slippery slope and often devs not familiar with RPM eg end up reinventing the wheel as each deployment failure breeds distrust between teams and more workarounds and safeguards get introduced.