I wonder if you could just madvise the entire address space. My hunch is that it's for performance reasons only - fewer pages to scan, hash, and de-duplicate.
1. You need a much better newer systemd than your distro likely packages.
2. KSM dedupe scans aren't free. Your system can spend its time doing that scan or it can spend its doing work with duplicate pages. Only relatively idle or highly homogenous systems would be free of the penalty.
3. For applications, especially statically linked ones, the duplicated code is not super likely to fall along page boundaries, thus the actual detectable duplication will be relatively low.
That said, it's still great for densely deploying a high traffic microservice that's more memory than CPU bound.