This means you actually have 4 Remote Runtime Environments (RRE), rather than 4 nodes in a single environment. This does mean you have to deploy to each RRE independently as you've pointed out. The benefits of this architecture, though, are that you can manage each of those completely independently - upgrading versions, applying hotfixes, etc. This is definitely what I recommend as a best practice. Even the fact that you CAN deploy decision independently to different nodes can be beneficial if you want to test a new decision with only a portion of your audience, for example.
If you want to streamline deployment across all RREs, I suggest looking in to the command line options for deploying decision campaigns so you can set something up to do it all with a single command.