That together with the screenshots points to NAS02 being in sync according to what it knows and NAS01 pulling stuff from NAS02. This pulling takes a long time and produces some sync conflicts (or a lot) and once there are enough conflicts, they are saved to the db and sent to NAS02. As NAS01 is stuck syncing (i.e. doesn’t scan), the “.DS_Store” issue does not resolve, but that doesn’t hurt. That the remote device completion is very low for NAS02 as displayed on NAS01 might also be due to conflicts: The conflicts detected in NAS01’s pulling iterations are sent, but if NAS01 “loses” the conflict they will only get resolved in the next puller iteration - and that is a long while off (when NAS01 is ± in sync for the first time). This might be optimized by checking for conflicts when computing remote device completions, but I don’t know whether it is worth the effort. I actually checked this and it is wrong: Winning conflicts are not wrongly reported as needed.
What remains unclear is your earlier statement that NAS01 has data that NAS02 does not: That should sync in any case as it should have been picked up in the scan before the current pull operation. You could grep the logs to check whether these unsynced files are mentioned at any point.
All in all I believe that syncing is on track, it just takes a lot of time as it is a lot of data.