I made some tests with volume quota but all seems pretty much OK. The quotas are stored to the disk and shared in the cluster and indicated by the settings you showed, but also (at least >= 7.4.2) the information is written down if the process is stopped/restarted, which is what happens in the cluster.
Were you running 7.4.1 before? In that case the issue may already be gone. If the issue is still present it might be a problem replicating the data through the cluster on shutdown. If the problem is still reproducible I recommend to file a Service Request with some instructions how to replicate the problem, so that support can give it a try and see if the issue is reproducible or collect the right information engineering needs to solve the problem.