You're doing a good job of backing up. I don't believe that most people are.
It's not something that would take us much time. We do sync to handhelds now and we can probably use that as the core for anything new.
Also since the Id is linux-based you have some pretty excellent FOSS incremental backup tools at your fingertips. Rsync, for example, is an open-source, standardized tool that can do remote incremental backups with a single shell command. It's super easy to script, very reliable, and has a very low network overhead. Additionally while it generally does incremental backups based on timestamp/filesize changes, it can do a more thorough, resource-intensive checksum-based incremental for a periodic "deep backup."
One major advantage of an rsync-type incremental backup is that it only sends the parts of the files that changed rather than resyncing the whole file when any part of it changes (which is what the current handheld sync implementation in JRiver does). That can be a huge time and bandwidth savings when you're talking about minor changes to files like tag changes or something. For example, if I write tags to my files, and redo genres on a bunch of music, the files themselves aren't much different (just the tag header), but JRiver's sync resyncs all of the files. An rsync backup would just send the chunks that changed, reducing bandwidth, transmission time, CPU use, etc. which might be handy if you're talking about automated user-transparent backups.
